WO2023063365A1 - Design execution device, design execution system, and design execution method - Google Patents

Abstract

This design execution device comprises: a design unit that includes, for respective functions, nodes including overview information prescribing information specifying processing for implementing functions without the substance of the aforementioned processing, the design unit enabling connection of a plurality of the nodes through an external operation, and generating execution order data prescribing an order of execution for a specific group of functions through connection of a specific group of nodes; a generation unit for generating configuration information whereby the overview information to use with each of the specific group of functions is prescribed in accordance with the order of execution, by inheritance of the overview information of a preceding-stage node among the specific group of nodes to a later-stage node connected to a later stage from the preceding-stage node in accordance with the execution order data generated by the design unit; and an execution unit for substantiating the processing for implementing the specific group of functions by using the configuration information generated by the generation unit, and executing the substantiated processing in the order of execution.

Description

DESIGN EXECUTION DEVICE, DESIGN EXECUTION SYSTEM, AND DESIGN EXECUTION METHOD Import by reference

This application is Japanese Patent Application No. 2021-168779 filed on October 14, 2021 and Japanese application filed on September 5, 2022. Two priority claims of Japanese Patent Application No. 2022-140736 are claimed, and the contents thereof are incorporated into the present application by reference.

The present invention relates to a design execution device, a design execution system, and a design execution method for designing and executing a target system.

As a background art in this technical field, Patent Document 1 discloses a program development support device that generates a program for executing data processing described in graph form on a target device. This program development support device includes a GUI section, a program generation section, a process execution function database, and a data transfer function database. In this program development support device, when a certain process included in data processing can be executed by a different kind of arithmetic unit installed in the target device, the processing execution function database stores the processing execution function database so that each arithmetic unit can execute the process. A data transfer function database holds a data transfer function for that purpose. The program development support device allows the GUI unit to select which processing unit is to execute the processing, and the program generation unit reads the processing execution function and the data transfer function corresponding to the selected processing unit, Generate a program to execute data processing on the target device.

JP 2017-062554 A

In order to generate a program for executing the target data processing on the target device, the technology described in Patent Document 1 needs to read the data processing method from the graph format. However, the details of the reading method are unclear.

In addition, even if the program is generated, the above-mentioned Patent Document 1 does not describe detailed processing for executing the program. must be prepared separately. Therefore, the behavior of the generated program cannot be verified.

The purpose of the present invention is to improve the convenience of design and execution.

A system design execution device, which is one aspect of the invention disclosed in the present application, has a node for each function that does not have the substance of a process that realizes the function, and that includes summary information that specifies information specifying the process, A design unit that can connect a plurality of the nodes and generates execution order data that defines the execution order of a specific function group by connecting a specific node group, and according to the execution order data generated by the design unit, The outline information used in each of the specified function groups is defined according to the execution order by inheriting the outline information of the preceding node of the specified node group to the succeeding node connected to the succeeding stage of the preceding node. a generation unit that generates configuration information; an execution unit that materializes processes for realizing the specific function group using the configuration information generated by the generation unit and executes the materialized processes in the execution order; characterized by having

According to the representative embodiment of the present invention, it is possible to improve the convenience of design and execution. Problems, configurations, and effects other than those described above will be clarified by the following description of the embodiments.

FIG. 1 is an explanatory diagram of a system configuration example of a design execution system according to a first embodiment. FIG. 2 is a block diagram showing a hardware configuration example of a computer. FIG. 3 is a block diagram of a functional configuration example of the design execution system according to the first embodiment; FIG. 4 is an explanatory diagram of a detailed system design method by the design department according to the first embodiment; FIG. 5 is an explanatory diagram of a detailed configuration information generation method by the generation unit according to the first embodiment; FIG. 6 is an explanatory diagram of detailed processing related to system execution of the function generation unit, generation unit, holding unit, execution unit, control unit, and processing device according to the first embodiment. FIG. 7 is a flowchart of a detailed processing procedure example from system design to execution by the design execution system according to the first embodiment. FIG. 8 is a flowchart of detailed processing procedures for ID assignment by the design execution system according to the second embodiment. FIG. 9 is an explanatory diagram of a detailed configuration information generation method by a generation unit according to the second embodiment; FIG. 10 is an explanatory diagram of an example of a detailed timing chart when the execution unit executes the system according to the second embodiment; FIG. 11 is a block diagram of a functional configuration example of a design execution system according to a third embodiment; FIG. 12 is an explanatory diagram of a detailed processing example of configuration information generation in the generation unit of the design execution system according to the fourth embodiment. FIG. 13 is an explanatory diagram showing a specific example of processing for duplication integration and node connection arrangement. FIG. 14 is a flowchart of detailed generation processing procedures by the generation unit of the design execution system according to the fourth embodiment. FIG. 15 is an explanatory diagram of a detailed processing example of configuration information generation in the generation unit of the design execution system according to the fifth embodiment. FIG. 16 is a flow chart showing detailed generation processing procedures by a generation unit of the design execution system according to the fifth embodiment; FIG. FIG. 17 is an explanatory diagram of a detailed processing example of configuration information generation in the generation unit of the design execution system according to the sixth embodiment. FIG. 13 is a flow chart showing detailed generation processing procedures by a generation unit of the design execution system according to the sixth embodiment; FIG. FIG. 19 is an explanatory diagram of a specific example 2 of the overlap integration and node connection arrangement processing according to the sixth embodiment. FIG. 20 is a flowchart of detailed generation processing procedures by the generation unit of the design execution system according to the sixth embodiment.

Hereinafter, embodiments according to the present invention will be described based on the drawings. The examples shown below are examples for explaining the present invention, and are appropriately omitted and simplified for clarity of explanation. The present invention can also be implemented in various other forms. Unless otherwise specified, each component may be singular or plural.

The position, size, shape, range, etc. of each component shown in the drawings may not represent the actual position, size, shape, range, etc. in order to facilitate the understanding of the invention. As such, the present invention is not necessarily limited to the locations, sizes, shapes, extents, etc., disclosed in the drawings.

As examples of various types of information, expressions such as "table", "list", and "queue" may be used for explanation, but various types of information may be expressed in data structures other than these. For example, various information such as "XX table", "XX list", and "XX queue" may be referred to as "XX information". When describing identification information, expressions such as “identification information”, “identifier”, “name”, “ID”, and “number” are used, but these can be replaced with each other.

When there are multiple components that have the same or similar functions, they may be described with the same reference numerals with different suffixes. Further, when there is no need to distinguish between these constituent elements, the subscripts may be omitted in the description.

In the examples, the processing performed by executing the program may be explained. Here, the computer executes a program by means of a processor (eg, CPU, GPU) and performs processing determined by the program while using storage resources (eg, memory) and interface devices (eg, communication port). Therefore, the main body of the processing performed by executing the program may be the processor. Similarly, a main body of processing executed by executing a program may be a controller having a processor, a device, a system, a computer, or a node. The subject of the processing performed by executing the program may be an arithmetic unit, and may include a dedicated circuit for performing specific processing. Here, the dedicated circuit is, for example, FPGA (Field Programmable Gate Array), ASIC (Application Specific Integrated Circuit), CPLD (Complex Programmable Logic Device), or the like.

The program may be installed on the computer from the program source. The program source may be, for example, a program distribution server or a computer-readable storage medium. When the program source is a program distribution server, the program distribution server may include a processor and storage resources for storing the distribution target program, and the processor of the program distribution server may distribute the distribution target program to other computers. Also, in the embodiment, two or more programs may be implemented as one program, and one program may be implemented as two or more programs.

<Design execution system>
The design execution system 100 shown in FIG. 1 has a server 111 and one or more clients 110 . The server and client are communicably connected via a network 101 such as the Internet, a LAN (Local Area Network), or a WAN (Wide Area Network). A server 111 is a computer that manages the client 110 .

The design execution system 100 has a function group creation function, a configuration information generation function, and a system execution function. The function group creation function creates functions required for system design for each processing unit by the operation of the user using the design execution system 100 . The configuration information generation function designs a system desired by an individual user by combining functions selected from a function group obtained from the function group creation function, and generates information on the designed system as configuration information. The system execution function executes the designed system from the configuration information.

The function group creation function, configuration information generation function, and system execution function may be implemented in either the server 111 or the client 110 as long as they are implemented in the design execution system 100 . For example, server 111 may implement the function group creation function, and client 110 may implement the configuration information generation function and the system execution function. Alternatively, the server 111 may implement the function group creation function and the system design function, and the client 110 may have the system execution function, receive the system designed by the server 111, and execute the received system.

A computer that implements the function group creation function is called a function group creation device, and a computer that implements at least the system execution function out of the configuration information generation function and the system execution function is called the design execution device. Also, in FIG. 1, the client-server type design execution system 100 is taken as an example, but a stand-alone type system design execution device may be used. In the first embodiment, for convenience of explanation, the design execution system 100, in which the server 111 is a function group creation device implementing a function group creation function and the client 110 is a design execution device implementing a system execution function, will be described as an example. do. Note that the client 110 may transmit information to a browser of an external device (which may be the server 111, not shown) that can communicate via the network 101, and be remotely operated on the browser of the external device.

<Computer hardware configuration example>
FIG. 2 is a block diagram showing a hardware configuration example of a computer (server 111, client 110). The computer 200 has a processor 201 , a storage device 202 , a processing device 203 , an input device 204 , an output device 205 and a communication interface (communication IF) 206 . Processor 201 , storage device 202 , processing device 203 , input device 204 , output device 205 and communication IF 206 are connected by bus 207 . Processor 201 controls computer 200 . A storage device 202 serves as a work area for the processor 201 . Also, the storage device 202 is a non-temporary or temporary recording medium that stores various programs and data. Examples of the storage device 202 include ROM (Read Only Memory), RAM (Random Access Memory), HDD (Hard Disk Drive), and flash memory.

The processing device 203 is an accelerator that speeds up specific functions, and specifically includes, for example, a GPU (Graphics Processing Unit), an AI (Artificial Intelligence) chip, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Array).

The input device 204 inputs data. Input devices 204 include, for example, various sensors such as a keyboard, mouse, touch panel, numeric keypad, scanner, and camera. The output device 205 outputs data. Output devices 205 include, for example, displays, printers, and speakers. Communication IF 206 connects to network 101 to transmit and receive data. Note that the processor 201 may be composed of a plurality of processors.

<Functional Configuration Example of Design Execution System 100>
FIG. 3 is a block diagram of a functional configuration example of the design execution system 100 according to the first embodiment. The server 111 has a function group creation unit 130 . The client 110 has a design unit 120 , a generation unit 121 , a storage unit 122 , an execution unit 123 , a control unit 124 and a processing device 203 .

Specifically, these are realized by causing the processor 201 to execute a program stored in the storage device 202 shown in FIG. 2 or by the storage device 202, for example. First, a functional configuration example on the server 111 side will be described.

As a function group creating function, the function group creation unit 130 creates a function group 420 necessary for system design by the user using the design execution system 100 and outputs the created function group 420 to the holding unit 122 . Each function of the function group 420 may be created in a program format such as a class or function, or may be created as a library for black-boxing the contents of processing.

Here, the processing device 203 can be selected for some of the functions of the function group 420, and the selected processing device 203 speeds up the processing. Henceforth, this function is described as a "specific high-speed function."

Next, a functional configuration example on the client 110 side will be described. As one of the configuration information generating functions, the design unit 120 designs a design target (system to be executed by the client 110) by connecting nodes having function overview information using a graphical user interface, and selects nodes adopted for the design. system information defining the execution order of the processes realized by the functions corresponding to is output to the generation unit 121 as execution order data. The function overview information does not have the substance of the processing that implements the function, but is information for identifying what kind of processing is to be performed. A numerical value defined within 100 may be used as the function overview information.

The generation unit 121 generates configuration information 320 that defines the function overview information according to the order of execution from the system information, and outputs the configuration information 320 to the holding unit 122 . Details of processing by the design unit 120 will be described later with reference to FIG. 4, and details of configuration information generation by the generation unit 121 will be described later with reference to FIG.

The holding unit 122 stores the function group 420 acquired from the function group creation unit 130 and the configuration information 320 acquired from the generation unit 121 and outputs them to the execution unit 123 . Here, if there is no change in the function group 420 or the configuration information 320, generation of the function group 420 or the configuration information 320 can be omitted. In this case, the holding unit 122 outputs the held function group 420 and the configuration information 320 to the execution unit 123 . This eliminates the need to regenerate the once-generated configuration information 320 and function group 420, thereby shortening the processing time and simplifying system startup after design.

The execution unit 123 materializes the system information described in the configuration information 320 acquired from the holding unit 122 for the corresponding functions in the function group 420, and executes the system. If the specified speed-up function is included in the materialized system, the execution unit 123 selects a device for processing according to the specified speed-up function, and executes the process in the processing device 203 via the control unit 124. .

The control unit 124 controls the processing device 203 to execute the specific speed-up function when the system materialized by the execution unit 123 includes the specific speed-up function.

The processing device 203 executes the processing of the specific speed-up function under the control of the control unit 124. Details of the execution unit 123, the control unit 124, and the processing device 203 will be described later with reference to FIG.

<Design Department 120>
FIG. 4 is an explanatory diagram showing a detailed system design method by the design unit 120. As shown in FIG. The design unit 120 can be accessed by entering a designated address from the web browser 400 and provides an operable interface. By enabling access and operation via the web browser 400 in this manner, access is possible from clients 110 other than the client 110 having the design unit 120, and remote design is also possible. However, the interface of the design section 120 is not limited to this. For example, it may be distributed as an installation-type tool and system design may be performed through a dedicated interface screen.

The design department 120 has a palette 20 that holds nodes used for system design, and a workspace 240 that is a workspace for designing the system, and displays them on the web browser of the client 110 .

The palette 20 is classified for each functional overview of the node. Specifically, for example, the category of input function 210 has nodes related to inputs of camera input 211 and video input 212 . A camera input 211 is a node that holds functional outline information regarding imaging of a camera that can be connected to the client 110 . The video input 212 is a node that holds function outline information for continuously reproducing images in units of frames for a captured moving image.

The process 220 has nodes related to human detection 221 and skeleton detection 222 . The person detection 221 is a node that holds function outline information for acquiring the position and size of a person from the acquired image and outputting coordinate information for enclosing the detected person in a rectangle. The skeleton detection 222 is a node that holds function overview information for detecting the position of a person's skeleton from an acquired image as coordinate information within the image.

The output function 230 has a result rendering 231 and an execution node 232 related to output. The result drawing 231 is a node that holds function overview information that the image obtained by the input function 210 is rendered with the result of execution by the process 220 . The execution node 232 is the last node connected to the system designed by the design unit 120 of the first embodiment, and the details of the operation of the execution node 232 will be described later with reference to FIG.

The above-described nodes held in the palette 20, the form of holding nodes for each function outline, and the function outlines to be classified are not limited to these, and are freely defined by the vendor who provides the design unit 120. , can be added and deleted.

By operating the input device 204, the user who designs the system arranges the necessary nodes from the palette 20 in the work space 240 and connects the nodes according to the flow of processing desired by the user. For example, when designing a system that detects the position of a person from an image captured by a camera, encloses the detected position in a rectangle, and displays it together with the image, the camera input 241, the person detection 242, and the result drawing 243 and an execution node 244 are arranged in the work space 240, and the nodes (camera input 241, human detection 242, result drawing 243, execution node 244) are connected by vectors 251 to 253, respectively, in the order of processing. .

Here, detailed settings for each node can be made for each node using detailed setting screens 246-248. The detailed setting screens 246 to 248 are displayed by specifying with the input device 204, for example. The detail setting screen 246 is a screen for setting the details of the function of the camera input 241, and sets the resolution and frame rate of the camera for shooting. The detailed setting screen 247 is a screen for setting the details of the function of the human detection 242, and sets a DNN (Deep Neural Network) used for the human detection 242 and a device for performing DNN calculation. The detail setting screen 248 is a screen for setting the details of the function of the result drawing 243, and sets a window size enlargement display function for the result drawing 243 and a processing speed display function of the designed system. The setting values in the detailed setting screens 246 to 248 can be changed by user operations.

As described above, system information is generated by the design unit 120 by connecting nodes on the workspace 240, and the generation unit 121 generates configuration information from this system information. In FIG. 4, for example, like the camera inputs 211 and 241, the nodes held in the palette 20 and the nodes arranged in the work space 240 are assigned different numbers even if the node names are the same. . This is different from the nodes held by the palette 20 in that the nodes arranged in the workspace 240 contain detailed setting values for use in the system, and the function outline information held by the nodes is also different. Different numbers are given for distinction.

<Configuration information generation example>
FIG. 5 is an explanatory diagram showing an example of generation of configuration information from system information by the generation unit 121. As shown in FIG. Each node (camera input 241, human detection 242, result drawing 243) that constitutes the system information is not an actual function, but function overview information 300, 301, 302. Note that the function overview information 300 , 301 , 302 also includes information set on the detail setting screens 246 , 247 , 248 .

The generation unit 121 outputs the function overview information 300, 301, and 302 of each node to the subsequent nodes. For example, camera input 241 outputs its functional summary information 300 to people detection 242 .

At this time, the latter node outputs integrated function overview information including the function overview information acquired from the previous node and the function overview information of the node to the latter node. For example, human detection 242 outputs integrated functional overview information 310 including functional overview information 300 obtained from camera input 241 and functional overview information 301 of human detection 242 to result rendering 243 . Also, the result rendering 243 renders the integrated function outline information 320 including the integrated function outline information 310 (function outline information 300, function outline information 301) acquired from the human detection 242 and the function outline information 302 of the result rendering 243 to the execution node. H.244 output.

Also, each node assigns an ID to the function overview information. As for the method of assigning IDs, for example, the ID of the function outline information 300 of the starting point node (camera input 241) that does not have a previous node is set to "0", and the function outline information received from the previous stage or the integrated function outline is assigned to the latter node. The maximum ID value described in the information is read, and a value obtained by incrementing the read ID value is assigned as the ID of the node. Therefore, the ID of the function outline information 301 of the human detection 242 is "1", and the ID of the function outline information 302 of the result drawing 243 is "2". However, the method of assigning IDs should be such that the nodes constituting the designed system do not have duplicate values, and the value of the ID of the node is not limited to the value incremented from the previous ID value.

The execution node 244 is a node connected to the end of the system designed by the design unit 120 of the first embodiment. Information 300-301 is included. The integrated function overview information 320 acquired by the execution node 244 becomes the configuration information generated by the generation unit 121 . Hereinafter, it will be referred to as configuration information 320, and the function overview information 300, 301, and 302 will be referred to as configurations 300, 301, and 302, respectively.

<Example of system execution>
FIG. 6 is an explanatory diagram showing detailed processing related to system execution of the function group creation unit 130, generation unit 121, holding unit 122, execution unit 123, control unit 124, and processing device 203. As shown in FIG.

The holding unit 122 holds the configuration information 302 generated by the generating unit 121 and the function group 420 generated by the function group generating unit 130 . The function group 420 holds the entities of functions (classes 421 to 423, for example) corresponding to all the nodes held in the palette 20 by the design unit 120. FIG.

The execution unit 123 selects and materializes the configurations 300, 301, and 302 described in the configuration information 320 from the function group 420 acquired from the holding unit 122. The execution unit 123 lists the materialized functions (camera input 241, human detection 242, result drawing 243) in the order of process execution, for example, like a list 431, and executes the listed processes in the execution base 432. . If the specified speed-up function is included in the materialized functions (camera input 241, human detection 242, result drawing 243), the execution unit 123 selects a device for processing according to the specified speed-up function, Processing is executed by the processing device 203 via the control unit 124 .

For example, in the configuration 301 of the human detection 242 in the configuration information 320, there is a description of ""device":"GPU"". According to this description, the execution unit 123 selects the GPU as the processing device 203 only for the processing of the human detection 242, and outputs the processing content and the information of the processing device 203 to the control unit 124 during the processing of the human detection 242. Based on the acquired information, the control unit 124 executes specified processing on the processing device 203 and outputs the processing result to the execution unit 123 .

<Example of Processing Procedure of Design Execution System 100>
FIG. 7 is a flowchart of a processing procedure example of the design execution system 100 according to the first embodiment. In function group creation, the server 111 distributes the function group created by the function group creating unit 130 to the client 110 (step S700).

In generating configuration information, the client 110 designs the system using a graphical user interface provided by the design unit 120 (step S720).

Next, the client 110 uses the generation unit 121 to generate configuration information from the designed system (step 521). The function group 420 generated by the server 111 and the configuration information 320 generated by the client 110 are retained in the client 110 by the retaining unit 122 (step S722). It is not necessary to hold the function group 420 and the configuration information 320 at the same time. For example, when the server 111 generates the function group 420 , only the function group 420 may be retained in the client 110 , and when the client 110 generates the configuration information 320 may be retained in the client 110 . The order of holding the function group 420 and holding the configuration information 320 is not limited.

In system execution, the client 110 reads the function group 420 and the configuration information 320 held by the holding unit 122 (step S740), and materializes the system information described in the configuration information 320 acquired by the execution unit 123 from the function group 420. (Step S742), the system is executed by the execution unit 123, the control unit 124, and the processing device 203 (Step S743).

For system execution, configuration information generation and function creation must be executed, and the function group 420 and the configuration information 320 must be held in the holding unit 122. 320, it is not necessary to execute configuration information generation and function creation each time the system is executed. That is, if the holding unit 122 holds the function group 420 and the configuration information 320, it is possible to execute only the system execution alone.

For example, during execution by the execution unit 123, there is no need to launch Nod-RED, which constitutes the design unit 120, and press the start button. Then, for example, the execution by the execution unit 123 is started only by turning on the power of the design execution device.

Also, configuration information generation and function generation may be executed only when there is a change in the configuration information 320 and the function group 420 respectively held by the holding unit 122 . In addition, by executing the processing of configuration information generation and system execution in a series of flows, the user can obtain an environment in which the system designed on the graphical interface can be immediately executed.

Thus, according to the first embodiment, the system is designed using a graphical user interface, the configuration information 320 is generated from the designed system, and the system is executed according to the generated configuration information 320. It is possible to execute the system immediately from the system design base that is used.

In particular, the process of generating the function group 420, the process of designing the system and generating the configuration information 320, and the process of executing the system from the generated configuration information 320 and the function group 420 are divided step by step. It is possible to execute the system execution in a series of flows, and it is also possible to execute the system execution separately, and it is possible to obtain a design execution system that simultaneously realizes the ease of system design and embedded implementation. .

The second embodiment will be described with a focus on the differences from the first embodiment. In addition, the same reference numerals are given to the points that are common to the first embodiment, and the description thereof will be omitted. In a second embodiment, a detailed processing method will be described in the case where the configuration information 320 to be generated includes parallel processing and multi-process processing.

FIG. 8 is a flowchart showing detailed processing procedures for ID assignment in the generation unit 121 of the design execution system 100 according to the second embodiment. The generation unit 121 determines whether the node to be processed is the start node (step S800). If there is no preceding node, it is determined to be the starting node. If it is the start node (step S800: Yes), the generation unit 121 assigns "0" to the ID of the node (step S801), and proceeds to the process of step S806.

If it is determined in step S800 that the node is not the starting node (step S800: No), the generation unit 121 acquires function overview information or integrated function overview information from the preceding node (step S802). .

If parallel processing or branching processing is included in the designed system information, the same processing of the same node may be assigned as a different ID in the integrated function summary information obtained from the preceding node, resulting in duplication.

For this reason, the generation unit 121 excludes the same process of the same node (step S803). Specifically, for example, when the same function overview information is included in a plurality of input function overview information or integrated function overview information, the generation unit 121 generates, for example, Leave only one and delete the others.

Also, if parallel processing or branch processing is included in the designed system information, the same ID may be assigned to the parallelized node even in the processing of another node in the acquired integrated function overview information.

For this reason, the generation unit 121 assigns a sub-ID for node identification to another process of another node to which the same ID is assigned (step S804). Specifically, for example, the generating unit 121 adds a unique branch number to the end of the ID value for the different node. Branch numbers are given in ascending order among identical IDs.

The generation unit 121 reads the maximum value of the ID from the integrated function overview information after deduplication in step S803 and sub-ID assignment to the same ID in step S804, and The incremented value is assigned as the ID of the node (step S805), and the process proceeds to step S806.

At step S806, the generation unit 121 confirms whether the node to be processed is the end node (step S806). Specifically, for example, if the node is an execution node, it is an end node. If it is an end node (step S806: Yes), the process proceeds to step S808, and if not (step S806: No), the process proceeds to step S807.

In step S807, the generation unit 121 outputs the function overview information processed by the node to the subsequent node, and returns to step S802.

In step S808, the generation unit 121 outputs the processed function overview information as the configuration information 320 to the holding unit 122, and the ID management of the node ends.

<Example of configuration information generation for system configuration including parallel processing>
FIG. 9 is an explanatory diagram of a detailed processing example of configuration information generation in the generation unit 121 of the design execution system 100 according to the second embodiment.

Since the camera input 241 is the starting point node, the generation unit 121 generates the function outline information 930 with the ID value set to "0", the vector 251 to the person detection 242 which is the subsequent node, and the vector 920 to the skeleton detection 910 . , the function overview information 930 is output.

In the human detection 242, the generation unit 121 reads the maximum ID value "0" from the acquired function overview information 930, and sets the incremented value "1" as the ID value of the node. The generation unit 121 generates the integrated function summary information 941 by adding the function summary information 931 of the node and the generated ID value “1” to the acquired function summary information 930, and generates the integrated function summary information 941 as the vector 252, which is the subsequent node. Outputs the function overview information 931 to the result drawing 243 .

In the skeleton detection 910, the generation unit 121 reads the maximum ID value "0" from the acquired function overview information 930, and sets the incremented value "1" as the ID value of the node. The generation unit 121 generates the integrated function outline information 942 by adding the function outline information 932 of the node and the generated ID value “1” to the acquired function outline information 930 , and generates the integrated function outline information 942 . 243 , the integrated function summary information 942 is output to the image distribution 911 in a vector 922 .

When the generation of the integrated function overview information 941 by the human detection 242 and the generation of the integrated function overview information 942 by the skeleton detection 910 are completed, the process of setting the ID value to "1" is performed by the human detection 242 and the skeleton detection 910. and exist.

In addition, since integrated function overview information 941 and integrated function overview information 942 are input to the result drawing 243 with two inputs, the function overview information 930 output by the camera input 241 overlaps in the result drawing 243 .

In the result drawing 243, the generation unit 121 eliminates redundant portions of the overlapping function outline information 930 of the camera input 241 based on the acquired integrated function outline information 941 and integrated function outline information 942 (step S803).

In the result drawing 243, the generation unit 121 further assigns sub-IDs for identification to the human detection 242 and skeleton detection 910 whose ID value is "1" (step S804). Sub-IDs are assigned by adding an underscore to the end of the existing ID. By this method, the value of ID for human detection 242 is set to "1_1" and the value of ID for skeleton detection 910 is set to "1_2". However, the method of assigning the sub-ID is not limited, and a sub-ID key may be generated and an identifiable ID may be assigned to the generated sub-ID.

In the result drawing 243, the generating unit 121 further reads the IDs excluding the sub-IDs of the function outline information that has been processed, and increments the value "2" from the maximum value "1" to the function of the node. This is the ID of the summary information 933 . The integrated function overview information 951 generated in this manner is output to the execution node 244, which is the subsequent node, in the vector 253. FIG.

In the image distribution 911, the generation unit 121 reads the maximum ID value "1" from the acquired integrated function overview information 942, and sets the incremented value "2" as the ID of the function overview information 934 of the node. . The generation unit 121 generates the integrated function overview information 952 by adding the function overview information 934 of the node and the generated ID value “2” to the acquired integrated function overview information 942, and generates the integrated function overview information 952, which is the subsequent node in the vector 923. It outputs integrated function summary information 952 to execution node 244 .

Here, ““type”:“multi-process”” included in the function overview information 934 of the node indicates that the process of image distribution 911 is assigned as a separate process, and the process is performed independently of other processes. It becomes the information to order A detailed processing method regarding multi-processing will be described later.

When the generation of the integrated function overview information 951 in the result drawing 243 and the generation of the integrated function overview information 952 in the image distribution 911 are completed, the process of setting the ID value to “2” is performed. and exist.

Also, since the execution node 244 having two inputs receives the integrated function outline information 951 and the integrated function outline information 952 , the execution node 244 outputs the function outline information 930 and the skeleton detection 910 output by the camera input 241 . overlaps with the function overview information 932 output by .

In the execution node 244 , the generation unit 121 eliminates redundant portions of the overlapping function outline information 930 of the camera input 241 based on the acquired integrated function outline information 951 and integrated function outline information 952 . In addition, since there is overlap in the skeleton detection 910 whose ID value excluding the sub ID is "1", the generation unit 121 eliminates the overlap. Here, the function overview information 932 of the skeleton detection 910 with "ID": "1" to which no sub-ID is assigned is deleted (step S803).

Also, the generation unit 121 assigns a sub-ID for identification to the function outline information 933 of the result drawing 243 and the function outline information 934 of the image distribution 911 whose ID value is "2". A sub-ID is assigned by adding an underscore to the existing ID and then assigning the sub-ID. By this method, the ID of the function outline information 933 of the result drawing 243 is assigned as "2_1", and the ID of the function outline information 934 of the image delivery 11 is assigned as "2_2".

Since the execution node 244 is the terminal node, the generating unit 121 generates the function overview information 960 that has been processed as configuration information (hereinafter referred to as configuration information 960) and outputs it to the holding unit 122.

<System execution example of configuration information including parallel processing>
FIG. 10 is an explanatory diagram showing an example of a detailed timing chart during system execution in the execution unit 123 of the design execution system 100 according to the second embodiment.

Upon receiving the completion of the processing of the camera input 1001, the execution unit 123 starts the processing of human detection 1002 and skeleton detection 1003 in parallel. For the parallel processing that starts processing at the same time, the next processing is started upon completion of all the parallel processing.

The execution unit 123 starts the result drawing 1004 upon receiving the completion of the human detection 1002 and the skeleton detection 1003, and when the result drawing 1004 ends, the camera input 1011 of the next frame is started. Here, for example, if there is no next image to be processed, the processing of the execution unit 123 ends.

The execution unit 123 starts the image distribution 1005 upon receiving the completion of the skeleton detection 1003 in the previous stage. Since the image distribution 1005 receives a multi-process designation, it is synchronized with the completion of the skeleton detection 1003 at the start of processing, but is executed asynchronously with other processing after the start of processing.

In this way, according to the second embodiment, it is possible to parallelize the processing by branching nodes and writing them in parallel when designing the system, thereby speeding up the processing of the entire system. Also, by designating a node with multiple processes when designing the system, it is possible to obtain the design execution system 100 capable of executing processes asynchronously with other processes.

The third embodiment will be described with a focus on the differences from the first and second embodiments. In addition, the same reference numerals are given to the points common to the first and second embodiments, and the description thereof is omitted. In Example 3, the function group creation unit 130 in the server 111 containerizes the function group 420 together with the execution base 432 and the program execution base 1101 using a container base such as docker.

In addition, the execution unit 123 holds the container execution base 1102 in order to execute processing that implements containerized functions. This makes it possible to process the function group 420 even if the server 111 and the client 110 have different processing environments such as the OS (Operating System) and the processor 201 .

In addition, by containerizing each function of the function group 420 provided by the server 111 for the plurality of clients 110, even if there is a difference in the OS, the processor 201, etc., the function group 420 can be individually arranged according to each client 110. Processing can be performed with a single function group 420 without customization.

FIG. 11 is a block diagram showing a functional configuration example of the design execution system 100 according to the third embodiment.

The function group creation unit 130 generates a function group 420 necessary for system design by a user using the design execution system 100 for each processing unit, and the generated function group 420 is stored in an execution base 432 using a container base such as docker. and the program execution base 1101 , and output to the holding unit 122 as a containerized function group 1120 .

The generating unit 121 generates configuration information 1103 from the designed system information and outputs it to the holding unit 122 .

The holding unit 122 outputs the acquired containerization function group 1120 and configuration information 1103 to the execution unit 123 .

The execution unit 123 has a containerization function group 1120 , a container execution base 1102 and configuration information 1103 . The execution unit 123 arranges the containerization function group 1120 acquired from the holding unit 122 in the upper layer of the container execution base 1102 and arranges the configuration information 1103 in the lower layer of the container execution base 1102 .

The containerized function group 1120 has a function group 420 , an execution base 432 and a program execution base 1101 . The program execution infrastructure 1101 for executing the function group 420 and the execution infrastructure 432 is different from the infrastructure for executing functions written in programming languages such as python and C++, and the OS installed in the client 110. It is the foundation of the guest OS necessary to execute functions on the OS.

As a result, the containerization function group 1120 can be executed without being affected by the hardware and software environment of the client 110. Note that when there is a change in the containerized function group 1120, the function group creating unit 130 reflects the change and newly arranges the containerized function group 1120. FIG.

The container execution base 1102 has the function of absorbing the difference between the environment of the client 110 and the environment of the containerized function group 1120 and executing the containerized function group 1120 regardless of the client 110 environment. The configuration information 1103 is generated by the same method as in the first and second embodiments and arranged in the hierarchy of the container execution infrastructure 1102 .

The execution unit 123 selects the configuration described in the configuration information 1103 from among the function groups 420 in the containerized function group 1120 acquired from the holding unit 122, and materializes each function. Then, the execution unit 123 executes processing using the program execution base 1101 and the container execution base 1102 by the execution base 432 . At this time, the execution unit 123 selects the processing device 203 for processing according to the speed-up function, and executes the processing in the processing device 203 via the control unit 124 . This speeds up the processing.

When outputting the containerization function group 1120 from the server 111 to the plurality of clients 110, for example, the containerization function group 1120 is connected to each client 110 through the container execution base 1102 through a port specific to each individual client. Appropriate processing for the client 110 may be required. In such a case, the holding unit 122 may automatically assign appropriate processing when connecting the containerized function group 1120 , or the design unit 120 may set a unique value for the client 110 .

As described above, according to the third embodiment, the function group 420 generated by the function group creation unit 130 in the server 111 is containerized and generated together with the execution base 432 and the program execution base using a container base such as docker. Also, the execution unit 123 holds the container execution base 1102 in order to execute the containerized function group 420 .

As a result, even if the server 111 and the client 110 have different processing environments such as the OS (Operating System) and the processor 201, the function group 420 can be processed. In addition, by containerizing the function group 420 provided by the server 111 for a plurality of clients 110, even if there is a difference in the OS, the processor 201, etc., the function group 420 does not have to be customized for each client 110. A single function group 420 can perform the processing.

Example 4 will be described with a focus on the differences from Example 1, Example 2, and Example 3. In addition, the same reference numerals are given to the points common to the first, second, and third embodiments, and the description thereof is omitted. In a fourth embodiment, a detailed processing method will be described in the case where the configuration information 320 to be generated includes parallel processing and multi-process processing.

However, unlike the second embodiment, each node cannot synchronize even if there are multiple inputs from the previous node, and when any input is received from the previous node, Assume a system that acquires the function overview information of the node in the previous stage, combines the function overview information of the node, and starts the process of outputting to the subsequent stage. Hereinafter, this system will be referred to as an asynchronous system.

In Examples 1 to 3, each node cannot synchronize even if there are multiple inputs from the previous node, so if the system to be designed includes parallel processing or multi-processing, A plurality of execution paths are generated, and an execution node is called for each execution path. For example, in the example shown in FIG. 9, execution node 244 is called three times because three execution paths are generated.

On the other hand, in order to consolidate the execution node calls by multiple execution paths into one call, it is conceivable to insert a synchronization node that aggregates multiple execution paths in the preceding stage of the execution node. However, setting the connection of each execution path to the synchronization node is a manual operation of the user. Therefore, if a connection setting error occurs due to a manual operation by the user, only a part of the execution path processes reaches the synchronization node, and the call to the execution node is not started. The user has to check whether the number of execution paths to be synchronized on the connection setting property screen is correct or not.

For this reason, the fourth embodiment provides a system that can execute synchronous processing without a synchronous node for parallel processing and multi-process processing in such an asynchronous system.

<Example of configuration information generation for system configuration including parallel processing>
FIG. 12 is an explanatory diagram of a detailed processing example of configuration information generation in the generation unit 121 of the design execution system 100 according to the fourth embodiment. In FIG. 12, there are a camera input 1200, an inference A 1201, an inference B 1202, an inference C 1204, a result drawing 1205, and an execution node 1206 as node groups.

The output of camera input 1200 is connected by vector 1210 to the input of inference A 1201 and by vector 1211 to the input of inference B 1202 . The output of Infer A 1201 is connected to the input of Render Result 1205 by vector 1212 .

The output of reasoning B 1202 is connected by vector 1213 to the input of reasoning A 1203 and by vector 1214 to the input of reasoning C 1204 . The output of Infer A 1203 is connected to the input of Render Result 1205 by vector 1215 . The output of Inference C 1204 is connected to the input of Render Result 1205 by vector 1216 . The output of result 1205 is connected to the input of execution node 1206 by vector 1217 .

A camera input 1200 is the starting node. The generation unit 121 sets the value of the ID to "AAA" and acquires "2" as the number of outputs, which is the number of nodes connected in the subsequent stage. Here, the ID "AAA" set by the generation unit 121 is arranged in system information (information that defines the execution order of processes realized by the functions corresponding to the nodes adopted in the design of the system executed by the client 110). is different from the ID given to any of the nodes.

This means setting different IDs for each node with different functions, as well as setting different IDs even when the same functions are arranged as different nodes. For example, inference A1201 and inference A1203 have the same function, but are arranged as different nodes, so different IDs are set for each.

The generation unit 121 sets the set ID "AAA", the number of outputs "2", and "camera input" indicating the function of the node in the camera input 1200 as the function outline information 1220. The generation unit 121 outputs the function overview information 1220 from the camera input 1200 to the subsequent inference A 1201 . The generation unit 121 also outputs the function overview information 1221 and the function overview information 1222 from the camera input 1200 to the reasoning B 1202 . In this case, since the output of the camera input 1200 is branched not only to the inference A 1201 but also to the inference B 1202, the number of outputs of the function overview information 1220 to 1222 is "2".

Note that the function overview information 1220 to 1222 are the same information as the function overview information. The reason why they are described separately is to explain that the asynchronous system outputs the function outline information separately by the number of branches. Details will be described later in the operation of each node.

Here, the motions of each of the subsequent nodes 1202 to 1206 that are not the starting point nodes other than the camera input 1200 will be described. Each of the nodes 1202 to 1206 sets a unique ID for the node, and outputs integrated function overview information including the function overview information obtained from the preceding node and the function overview information for the node to the subsequent node.

Here, the function overview information includes the number of outputs, but if the output of the subsequent node does not branch, the number of outputs obtained from the preceding node is inherited and output to the subsequent node. For example, although the number of outputs from a preceding node of a certain node is "2", the output of the succeeding stage does not branch, that is, there is one succeeding node. In this case, the number of outputs "2" is inherited by the succeeding node. This is because the number of outputs "2" has already been counted as "1" for the output to the succeeding node (the route from the start node to the end node via).

On the other hand, if the output of the subsequent node is branched, the number of outputs is updated according to the obtained number of outputs and the number of branches of the subsequent node, and is output to the subsequent node. For example, if the number of outputs obtained from the preceding node is "2" and the output of the subsequent stage branches into two, that is, if the number of subsequent nodes is two, the path newly branched to the number of outputs "2" "1" for the number of outputs is added to make the number of outputs "3". This output number "3" is output to each of the two succeeding nodes.

Each node cannot synchronize at the node even if there are multiple inputs from the previous node, and when there is any input from the previous node, the function overview information of the previous node is It acquires and starts the process of outputting it to the subsequent stage together with the function outline information of the node.

The generation unit 121 sets the ID value to "BBB" in the inference A1201, and acquires the function overview information 1220 from the camera input 1200, which is the preceding node. In addition, since the node connected to the subsequent stage of inference A1201 is only result drawing 1205, and the output of inference A1201 does not branch, the number of outputs "2" obtained from camera input 1200 is inherited, and the number of outputs of inference A1201 is becomes "2". That is, in the number of outputs "2", "1" for 1212 vectors indicating the output (the path from the camera input 1200 to the end node via the inference A 1201) to the result rendering 1205, which is the subsequent node of the inference A 1201, already exists. This is because it is counted at 1200.

In the inference A 1201, the IDs "BBB", "inference A", and the number of outputs "2" are used as the function outline information 1223. The generation unit 121 outputs the integrated outline information including the function outline information 1223 of the reasoning A 1201 and the function outline information 1220 acquired from the camera input 1200 to the subsequent result drawing 1205 .

The generation unit 121 sets the ID value to "CCC" in inference B1202, and acquires the function overview information 1221 from the camera input 1200, which is the preceding node. Since the node following the inference B 1202 branches to the inference A 1203 and the inference C 1204, the number of outputs of the inference B 1202 is "2", which is obtained from the camera input 1200. 3”.

That is, in the number of outputs "2", "1" for 1213 vectors indicating the output (the path from the camera input 1200 to the end node via the camera input 1200) to the inference A 1203, which is one subsequent node of the inference B 1202, already exists in the camera Although it is counted at the input 1200, "1" for 1214 vectors indicating the output (the path from the camera input 1200 to the end node via) to the inference C 1204, which is the other subsequent node of the inference B 1202, is counted. because there is no

In the inference B 1202, the IDs "CCC", "inference B", and the number of outputs "3" are used as the function outline information 1224. The generation unit 121 outputs integrated overview information including the function overview information 1224 of the inference B 1202 and the function overview information 1221 acquired from the camera input 1200 to the inference A 1203 and inference A 1204 in the subsequent stage.

The generation unit 121 sets the ID value to "BBB'" in the inference A1203, and acquires the function overview information 1221 and 1224 from the inference B1202, which is the preceding node. Since the only node connected to the subsequent stage of reasoning A1203 is result rendering 1205, and the output of reasoning A1203 does not branch, the number of outputs "3" obtained from reasoning B1202 is inherited, and the number of outputs of reasoning A1203 is "3". becomes.

In the inference A 1203, the ID "BBB'", the "inference A", and the number of outputs "3" are used as the function outline information 1229. The generation unit 121 outputs the integrated overview information including the function overview information 1229 of the reasoning A 1203 and the function overview information 1221 and 1224 acquired from the reasoning B 1202 to the subsequent result drawing 1205 .

The generation unit 121 sets the ID value to "EEE" in the inference C1204, and acquires the function overview information from the inference B1202, which is the preceding node. Since the only node connected to the subsequent stage of inference C1204 is result drawing 1205, and the output of inference C1204 does not branch, the number of outputs "3" obtained from inference B1202 is inherited, and the number of outputs of inference C1204 is "3". becomes.

In the inference C 1204, the ID "EEE", the "inference C", and the number of outputs "3" are set as the function overview information 1230. The generation unit 121 outputs the integrated overview information including the function overview information 1230 of the reasoning C 1204 and the function overview information 1221 and 1224 acquired from the reasoning B 1202 to the subsequent result drawing 1205 .

The result rendering 1205 has three nodes in the preceding stage: inference A1201, inference A1203, and inference C1204. Since the fourth embodiment assumes an asynchronous system, the result rendering 1205 starts when the output of one of the preceding nodes 1201, 1203, and 1204 is input to the result rendering 1205. FIG. In other words, the result drawing 1205 receives outputs from the three preceding nodes 1201, 1203, and 1204 and operates three times.

Regarding the result drawing 1205, the operation when receiving the output from the reasoning A 1201 will be explained. The generation unit 121 sets the ID value to “DDD” in the result drawing 1205, and acquires the function overview information 1220 and 1223 from the inference A 1201, which is the preceding node. Execution node 1206 is the only node connected to the subsequent stage of result rendering 1205, and the output of result rendering 1205 does not branch. becomes "2".

In the result drawing 1205, the ID "DDD", "result drawing", and the number of outputs "2" are set as the function outline information 1226. The generation unit 121 outputs integrated overview information 1231 including the function overview information 1226 of the result drawing 1205 and the function overview information 1220 and 1223 acquired from the reasoning A 1201 to the subsequent execution node 1206 .

Regarding the result drawing 1205, the operation when receiving the output from the reasoning A 1203 will be explained. The generation unit 121 sets the value of ID to "DDD" in the result drawing 1205, and acquires function outline information 1221, 1224, and 1229 from the inference A 1203, which is the preceding node. Execution node 1206 is the only node connected to the subsequent stage of result rendering 1205, and the output of result rendering 1205 does not branch. becomes "3".

In the result rendering 1205, the ID "DDD", "result rendering", and the number of outputs "3" are set as the function outline information 1227. The generation unit 121 outputs integrated overview information 1232 including the function overview information 1227 and the function overview information 1221, 1224, and 1229 acquired from the inference A 1203 to the subsequent execution node 1206. FIG.

Regarding the result drawing 1205, the operation when receiving the output from the inference C 1204 will be explained. The generation unit 121 sets the ID value to “DDD” in the result drawing 1205, and acquires the function outline information 1222, 1225, and 1230 from the inference C 1204, which is the preceding node. Execution node 1206 is the only node connected to the subsequent stage of result rendering 1205, and the output of result rendering 1205 does not branch. becomes "3".

In the drawing result 1205, the ID "DDD", "drawing result", and the number of outputs "3" are set as the function overview information 1228. The generation unit 121 outputs integrated overview information 1233 including the function overview information 1228 and the function overview information 1222, 1225, and 1230 acquired from the inference C 1204 to the subsequent execution node 1206. FIG.

In this way, in an asynchronous system, there is a possibility that some functional summary information will be input redundantly at the aggregated node. Note that the order of outputting the three nodes of reasoning A1201, reasoning A1203, and reasoning C1204 to the result drawing 1205 is not limited, and the order may be changed.

The preceding node of the execution node 1206 is the result drawing 1205, but since the number of inputs of the result drawing 1205 is "3", the result drawing 1205 is executed three times. Since the result drawing 1205 outputs each execution result to the execution node 1206, the execution node 1206 is also executed three times.

The execution node 1206 asynchronously acquires the integrated overview information 1231, the integrated overview information 1232, and the integrated overview information 1233 in random order. The three pieces of integrated outline information 1231-1233 include the same function outline information redundantly, such as function outline information 1220-1222 related to the camera input 1200, for example. The detailed processing of the execution node 1206 for performing the synchronization processing after deduplication of the same information for the integrated summary information in which the duplicated information is divided and output multiple times by the asynchronous system will be described.

The execution node 1206 has function outline information acquisition 1240 , output branch number determination 1241 , overlap integration 1242 , node connection arrangement 1243 , and configuration information generation 1244 .

The function overview information acquisition 1240 acquires the integrated overview information 1231, the integrated overview information 1232, and the integrated overview information 1233 output from the result drawing 1205 in the previous stage in random order, and outputs them to the output branch number determination 1241.

The output branch count determination 1241 counts the number of acquired integrated overview information for each of the integrated overview information 1231 to 1233 acquired in random order. Also, the output branch number determination 1241 compares the counted number of integrated overview information with the number of outputs held in each of the integrated overview information 1231-1233.

The output branch count determination 1241 accumulates the acquired integrated overview information until the counted number of integrated overview information matches the number of outputs. When the counted number of pieces of integrated outline information and the number of outputs match, the output branch number judgment 1241 outputs the accumulated integrated outline information to the overlapping integration 1242 .

The operation of the output branch number determination 1241 will be described with specific examples. First, when the output branch number determination 1241 acquires the integrated outline information 1231, the output branch number determination 1241 counts the number of acquired integrated outline information (initial value is "0") and sets "1". , and confirms that the output number of the integrated overview information 1231 is “2”, and accumulates the integrated overview information 1231 .

Next, when the output branch number determination 1241 acquires the integrated outline information 1232, the output branch number determination 1241 counts the number of acquired integrated outline information and sets it to "2", and the output number of the integrated outline information 1232 is "2". 3”, and accumulates the integrated summary information 1232 .

Finally, when the output branch number determination 1241 acquires the integrated outline information 1233, the output branch number determination 1241 counts the number of acquired integrated outline information and sets it to "3", and the output number of the integrated outline information 1233 is After confirming that it is "3", the integrated summary information 1233 is accumulated. The output branch number determination 1241 duplicates the integrated overview information 1231, the integrated overview information 1232, and the integrated overview information 1233 accumulated so far because the counted acquisition number "3" and the output number "3" match. Output to integration 1242 .

The duplication integration 1242 integrates duplication of node functions using the IDs included in all of the acquired integrated overview information, and outputs the integrated integrated overview information to the node connection organizing 1243 as duplicate integrated overview information.

The node connection arrangement 1243 confirms the connection relationship of each node with respect to the overlap integration overview information acquired from the overlap integration 1242, and for a node having multiple preceding nodes, when output of all preceding nodes to the node is completed, Synchronous processing information for starting processing is added, and output to the configuration information generation 1244 as synchronous duplication integration summary information. A connection relation between nodes means a relation in which nodes are connected by a vector. Here, the processing of duplication integration 1242 and node connection arrangement 1243 will be specifically described with reference to the drawings.

FIG. 13 is an explanatory diagram showing a specific example of the processes of duplication integration 1242 and node connection arrangement 1243 according to the fourth embodiment. FIG. 13 shows changes in the integrated outline information when the overlap integration 1242 and the node connection arrangement 1243 in the execution node 1206 generate the synchronous overlap integrated outline information group 1302 from the acquired integrated outline information group 1300 .

The integrated overview information group 1300 includes integrated overview information 1231 to 1233 shown in FIG. In addition, in each of the integrated summary information 1231 to 1233, there is a connection relationship between left and right adjacent blocks. Taking the integrated overview information 1231 as an example, there is a node connection relationship with the function overview information 1220 and 1223 and a node connection relationship with the function overview information 1223 and 1226 .

The duplication integration 1242 determines and integrates function outline information having the same ID as the same node. For example, the function outline information 1220 to 1222 of the camera input 1200 are given the same ID "AAA", so the overlap integration 1242 determines that they are the same node and integrates them as overlap integration outline information 1311. Output.

On the other hand, the function summary information 1223 of the reasoning A 1201 and the function summary information 1229 of the reasoning A 1203 have the same function (reasoning A), but have different IDs (“BBB” and “BBB′”). , are not integrated by judging that they are different nodes.

In addition, since the function overview information 1224 of the reasoning B 1202 and the function overview information 1225 of the reasoning B 1202 are given the same ID "CCC", the overlapping integration 1242 determines that they are the same node and integrates them. Output as summary information 1312 . Since the function outline information 1226 to 1228 of the result rendering 1205 are given the same ID “DDD”, the overlap integration 1242 determines that they are the same node, integrates them, and outputs them as overlap integration outline information 1313 . After all of the duplicated function outline information have been integrated with the acquired integrated outline information, the duplicated integrated outline information 1311 to 1313 and the function outline information 1223, 1229, and 1230 are grouped as a duplicated composite outline information group 1301, and node connections are arranged. Output to 1243.

The node connection arrangement 1243 confirms the connection relationship of each node with respect to the overlap integration summary information group 1301 acquired from the overlap integration 1242, and for a node having a plurality of preceding nodes, all the preceding nodes output to the node. Synchronous processing information for starting the processing upon completion is added, and output to the configuration information generation 1244 as synchronous duplication integration summary information.

Specifically, for example, in overlapping composite overview information 1311 to 1313 and function overview information 1223, 1229, and 1230 in a group of overlapping composite overview information 1301, the connection relationships of nodes in horizontally adjacent blocks are Inherited from group 1300 .

For example, taking the overlapped composite overview information 1311 as an example, the node connection relationship between the function overview information 1220 and 1223 is inherited for the node connection relationship with the function overview information 1223 . In addition, as for the connection relationship with the overlapping composite overview information 1312, the connection relationship of the nodes between the function overview information 1221 and 1224 and the connection relationship of the nodes between the function overview information 1222 and 1225 are inherited.

On the other hand, the node connection arrangement 1243 confirms whether or not there is a node connection relationship between the function overview information 1223 and the duplicate/complex overview information 1313 . The function overview information 1223 has a node connection relationship with the function overview information 1226 . The function overview information 1226 is integrated with the overlapping integrated overview information 1313 . Therefore, the node connection sorting out 1243 determines that there is a node connection relationship between the function overview information 1223 and the duplicate/complex overview information 1313 . Therefore, the node connection arrangement 1243 connects the function overview information 1223 and the overlapping compound overview information 1313 . In this way, the node connection arrangement 1243 confirms the connection relationship of the nodes in the group of overlapping integrated overview information 1301 .

Also, the preceding node of the function overview information 1226 of the result drawing 1205 is the function overview information 1223 of the inference A1201. The preceding node of the function outline information 1227 of the result rendering 1205 is the function outline information 1229 of the reasoning A 1203 . The preceding node of the function outline information 1228 of the result drawing 1205 is the function outline information 1230 of the inference C1204. Since the same ID "DDD" is assigned to the function overview information 1226 to 1228 of the result rendering 1205, they are the same node. Therefore, it is determined that the result drawing 1205 has three nodes (the function outline information 1223 of the inference A 1201, the function outline information 1229 of the inference A 1203, and the function outline information 1230 of the inference C 1204) in the preceding node.

The node connection arrangement 1243 starts processing after acquiring the outputs of the three preceding nodes (the function outline information 1223 of the reasoning A 1201, the function outline information 1229 of the reasoning A 1203, and the function outline information 1230 of the reasoning C 1204) for the result drawing 1205. For this purpose, the synchronous processing information is added to the group 1301 of overlapping/composite summary information to form a group 1302 of synchronous deduplication integrated summary information, which is output to the configuration information generation 1244 .

Synchronous processing information is information that starts processing when output to all of the preceding nodes that exist in multiple nodes is completed. In this example, since there are three nodes (function summary information 1223 of reasoning A 1201, function summary information 1229 of reasoning A 1203, and function summary information 1230 of reasoning C 1204) in the preceding stage of result rendering 1205, overlap integration indicating result rendering 1205 Synchronization processing information 1320 is added to the overview information 1313 . This synchronous processing information 1320 has a function of causing the result drawing 1205 to start processing when output is received from all of the function outline information 1223 , 1229 , and 1230 .

As a result, the result drawing 1826 waits for the outputs of the inference A1823, inference A1829, and inference C1830 to start processing and synchronize the processing. Note that although there is no preceding node in the start node, for example, since the overlapped composite summary information 1311 is camera input 1200, camera input 1200 is started when an interrupt occurs in all of a plurality of nodes other than camera input 1200. It is possible to add synchronous processing information such as

In this way, the node connection arrangement 1243 completes the node connection confirmation, generates the synchronous duplication integration summary information group 1302 to which the synchronization processing information 1320 is added, and outputs it to the configuration information generation 1244 .

The configuration information generation 1244 generates the synchronous duplication integration summary information group 1302 acquired from the node connection arrangement 1243 as the configuration information 320 .

<Example of processing procedure of generation unit 121>
FIG. 14 is a flowchart of detailed generation processing procedures by the generation unit 121 of the design execution system 100 according to the fourth embodiment. The generation unit 121 determines whether the node to be processed is the start node (step S1400). If there is no preceding node, it is determined to be the starting node. If it is not the start node (step S1400: No), the generation unit 121 acquires the function outline information from the previous node (step S1401), and proceeds to the process of step S1402. If it is determined in step S1400 that the node is the start node (step S1400: Yes), the process proceeds to step S1402.

In step S1402, the generating unit 121 sets a node-specific ID for the node, and proceeds to step S1403.

In step S1402, the generation unit 121 acquires the number of outputs from the number of nodes connected in the subsequent stage, and proceeds to the processing of step S1404.

In step S1404, the generation unit 121 outputs integrated function overview information including a unique ID, the number of outputs, and the function overview information of the node to the function overview information acquired from the preceding node to the subsequent node. The process proceeds to S1405.

At step S1405, the generation unit 121 confirms whether the node to be processed is the end node (step S1405). Specifically, for example, if the node is an execution node, it is an end node. If it is an end node (step S1405: Yes), the process proceeds to step S1406, and if not (step S1405: No), the process proceeds to step S1400.

In step S1406, the generation unit 121 acquires integrated overview information in function overview information acquisition 1240, and proceeds to the processing of step S1407.

In step S1407, the generation unit 121 counts the number of acquisitions of the integrated overview information in the output branch count determination 1241, compares it with the number of outputs held in each integrated overview information, and if they match (step S1407: Yes), the accumulated integrated overview information is output to the duplication integration 1242, and the process proceeds to step S1408.

In step S1408, the generation unit 121 integrates duplicate functions of the nodes using the IDs included in all the acquired integrated outline information in the overlap integration 1242, and converts the integrated integrated outline information into duplicate integrated outline information. generated, and proceeds to step S1409.

In step S1409, the generation unit 121 confirms the connection relationship of each node from the acquired overlap integration summary information in the node connection arrangement 1243, and for a node having a plurality of preceding nodes, Synchronous processing information is added so as to start processing upon completion of output to the synchronous duplication integration summary information group 1301, and the process advances to step S1410.

In step S1410, the generation unit 121 generates the acquired synchronous overlap integration summary information group 1301 as the configuration information 320 in the configuration information generation 1244, and the processing of the generation unit 121 ends.

As described above, according to the fourth embodiment, in the asynchronous system, each node in the design unit 120 cannot be synchronized even when there are multiple inputs from the preceding node. Also, when the asynchronous system receives any input from the preceding node, it acquires the function summary information of the preceding node, combines the function summary information of the node, and starts the process of outputting it to the subsequent stage. put away. Even in such an asynchronous system, synchronous processing can be executed as a parallel system that is designed to branch.

Example 5 will be described with a focus on the differences from Examples 1 to 4. The same reference numerals are given to the points common to the first to fourth embodiments, and the description thereof will be omitted. In the fifth embodiment, a detailed processing method will be described in a case where the generated configuration information 320 includes parallel processing and multi-process processing in an asynchronous system.

However, unlike the fourth embodiment, assume a system in which the nodes branched for parallelization are aggregated into one node once, parallelization is aggregated, and then parallelization is performed again. The fifth embodiment provides a system capable of executing synchronous processing for parallel processing and multi-process processing in such a system.

<Example of configuration information generation for system configuration including parallel processing>
FIG. 15 is an explanatory diagram of a detailed processing example of configuration information generation in the generation unit 121 of the design execution system 100 according to the fifth embodiment. In FIG. 15, a camera input 1500, an inference A 1501, an inference B 1502, an inference C 1504, a result drawing 1505, an image distribution 1506, and an execution node 1507 exist as a node group.

The output of camera input 1500 is connected by vector 1510 to the input of inference A 1501 and by vector 1511 to the input of inference B 1502 . The output of reason A 1501 is connected by vector 1512 to the input of reason C 1504 .

The output of Inference B 1502 is connected by a vector 1513 to the input of Inference C 1504. The output of Inference C 1504 is connected by vector 1510 to the input of Render Result 1505 , and the output of Inference C 1504 is connected by vector 1511 to the input of Image Delivery 1506 .

The output of render result 1505 is connected to the input of execution node 1507 by vector 1517 . The output of image distribution 1506 is connected to the input of execution node 1507 by vector 1513 .

The fifth embodiment differs from the fourth embodiment in that flag signals are added to branched nodes after parallelization. A flag signal is a signal for distinguishing a branch.

The generation unit 121 sets the value of the ID to "AAA" in the camera input 1500, and sets the set ID "AAA" as the function outline information 1520. The generation unit 121 outputs the function overview information 1520 from the camera input 1500 to the inference A 1501 and the inference B 1502 in the subsequent stage. In this case, since the output of the camera input 1500 is branched not only to the inference A 1501 but also to the inference B 1502, the number of outputs is "2".

Furthermore, in Example 5, the generation unit 121 outputs a flag signal for a node that has a plurality of nodes in the subsequent stage and has a branch. Specifically, in the camera input 1500, the generation unit 121 outputs the flag signal “1” and the flag signal “2” together with the function overview information 1520 to the inference A 1501 and the inference B 1502 in the subsequent stages.

Note that when there are a plurality of post-stage nodes, when outputting the function overview information to the post-stage nodes, the generation unit 121 assigns individual numbers to the same information, as in the fourth embodiment. Function outline information may be given to the subsequent node. However, in the fifth embodiment, in order to emphasize the difference from the fourth embodiment, some branches will be given the same numbers and the function overview information will be output to the subsequent node.

For this reason, the function overview information of the camera input 1500 may be assigned different numbers and output to the reasoning A 1501 and the reasoning B 1502 as in the fourth embodiment, but here the function overview information 1520 assigned the same number is , the generation unit 121 outputs to inference A 1501 and inference B 1502 .

The generation unit 121 sets the ID value to "BBB" in the inference A1501, and acquires the function overview information 1520 and the flag signal "1" from the camera input 1500, which is the preceding node. Inference C1504 is the only node connected to the stage after Inference A1501, and the output of Inference A1501 does not branch. becomes.

In the reasoning A 1501, the function outline information 1521 is the ID "BBB", the "reasoning A", the number of outputs "2", and the flag signal "1". The generation unit 121 outputs integrated overview information including the function overview information 1521 and the function overview information 1520 acquired from the camera input 1500 to the inference C1504 in the subsequent stage.

The generation unit 121 sets the ID value to "CCC" in inference B1502, and acquires the function overview information 1520 from the camera input 1500, which is the preceding node. Inference C1504 is the only node connected to the stage after Inference B1502, and the output of Inference B1502 does not branch. becomes.

In the inference B 1502, the ID "CCC", the "inference B", the number of outputs "2", and the flag signal "2" are used as the function overview information 1522. The generation unit 121 outputs integrated overview information including the function overview information 1522 and the function overview information 1520 acquired from the camera input 1500 to the inference C1504 in the subsequent stage.

The generation unit 121 sets the ID value to "DDD" in the inference C1504, and acquires the function overview information from the inference A1501 and inference A1502, which are the preceding nodes, in random order. As the operation of the inference C 1504, first, the operation when receiving the function overview information 1521 from the inference A 1501 will be described.

The generation unit 121 sets the ID value to "DDD" in the inference C1504, and acquires the function overview information 1520 and 1521 from the inference A1501, which is the preceding node. The generation unit 121 acquires the number of outputs “2” from the number of nodes (result rendering 1505, image distribution 1506) connected in the subsequent stage by reasoning C1504.

However, the generation unit 121 detects the presence of the flag signal "1" by inference C1504, and detects that there is already a branch at the previous node including the previous node. Then, the generation unit 121 sets the number of outputs to "4" in order to consider all branch combinations considering the branches of the nodes before the previous stage and the branches of the nodes of the subsequent stage. That is, the number of flag signals is set to "2"×the number of outputs is "2", and the number of outputs is set to "4".

Also, the generation unit 121 prepares the flag signal "3" and the flag signal "4" because the node after the inference C1504 is branched. Then, in inference C 1504 , ID “DDD” and “inference C” are set as function overview information 1523 . The generation unit 121 outputs the function outline information 1523 and the function outline information 1520 and 1521 acquired from the preceding nodes as integrated outline information to the subsequent result rendering 1505 and image distribution 1506 . The generation unit 121 also outputs the flag signal “3” and the flag signal “4” prepared by the inference C 1504 to the result drawing 1505 and the image distribution 1506, respectively.

Secondly, as the operation of the inference C 1504, the operation when receiving the function overview information 1522 from the inference B 1502 will be described.

The generation unit 121 sets the ID value to "DDD" in the inference C1504, and acquires the function overview information 1520 and 1522 from the inference B1502, which is the preceding node. The generation unit 121 acquires the number of outputs “2” from the number of nodes (result rendering 1505, image distribution 1506) connected in the subsequent stage by reasoning C1504.

However, the generation unit 121 detects the presence of the flag signal "1" by inference C1504, and detects that there is already a branch at the previous node including the previous node. Then, the generation unit 121 sets the number of outputs to "4" in order to consider all branch combinations considering the branches of the nodes before the previous stage and the branches of the nodes of the subsequent stage.

Also, the generation unit 121 prepares the flag signal "3" and the flag signal "4" because the node after the inference C1504 is branched. Then, in inference C 1504 , the ID “DDD” and “inference C” are used as function overview information 1524 . The generation unit 121 outputs the function outline information 1524 and the function outline information 1520 and 1522 acquired from the preceding nodes as integrated outline information to the subsequent result rendering 1505 and image distribution 1506 . The generation unit 121 also outputs the flag signal “3” and the flag signal “4” prepared by the inference C 1504 to the result drawing 1505 and the image distribution 1506, respectively.

Next, the processing of result rendering 1505 and image distribution 1506 will be described. Inference C 1504 precedes result rendering 1505 and image delivery 1506 . The processing of result drawing 1505 and image distribution 1506 will first be described by focusing only on the processing of inference C 1504 , result drawing 1505 and image distribution 1506 which are preceding nodes.

The generation unit 121 sets the ID value to "EEE" in the result rendering 1505, and acquires the function outline information 1520, 1521, 1523 and the flag signal "3" from the inference C1504, which is the preceding node. Execution node 1507 is the only node connected to the subsequent stage of result rendering 1505, and the output of result rendering 1505 does not branch. It becomes "4".

In the drawing result 1505, the ID "EEE", "drawing result", the number of outputs "4", and the flag signal "3" are used as function overview information 1525. The generation unit 121 outputs integrated overview information including the function overview information 1525 and the function overview information 1520, 1521, and 1523 acquired from the inference C 1504 to the subsequent execution node 1507. FIG.

The generation unit 121 sets the ID value to "FFF" in the image delivery 1506, and acquires the function overview information 1520, 1521, and 1523 and the flag signal "4" from the inference C1504, which is the preceding node. Since the execution node 1507 is the only node connected to the subsequent stage of the image delivery 1506, and the output of the image delivery 1506 does not branch, the number of outputs "4" acquired from the inference C1504 is inherited, and the number of outputs of the image delivery 1506 is It becomes "4".

In the image delivery 1506, the ID "FFF", "image delivery", the number of outputs "4", and the flag signal "4" are used as function overview information 1526. The generation unit 121 outputs the integrated overview information including the function overview information 1526 and the function overview information 1520, 1521, and 1523 acquired from the inference C 1504 to the subsequent execution node 1507. FIG.

Regarding the processing of result rendering 1505 and image distribution 1506, secondly, the processing including branching of nodes prior to inference C1504, which is the preceding node, will be described.

Focusing only on the processing of the inference C 1504, which is the preceding node of the result rendering 1505, the result rendering 1505, and the image distribution 1506, as described above, the output from the inference C 1504 is transferred to the result rendering 1505 and the image distribution 1506 respectively. obtained and processed.

However, since there is a branch in the path passing through vector 1510 and vector 1511 at the node before the preceding node (inference C1504), in inference C1504, two processes are started according to the outputs from inference A1501 and inference B1502. . That is, the inference C 1504 outputs the function overview information 1523 following the path of the inference A 1501 and the function overview information 1524 following the path of the inference B 1502 to the result drawing 1505 and the image delivery 1506, respectively.

For this reason, the execution node 1507 has integrated overview information including the function overview information 1520, 1521, 1523, 1525; , 1527 and integrated overview information including function overview information 1520, 1522, 1524, 1528 are obtained. That is, execution node 1507 is activated four times.

Regarding the detailed processing of the execution node 1507, only differences from the fourth embodiment will be explained. Execution node 1507 differs from execution node 1206 of the fourth embodiment in that flag signal confirmation 1540 is added.

If there are branches before the previous node, it is necessary to generate configuration information considering all branches. Therefore, in order to confirm whether the execution node 1507 contains the function overview information including all nodes, a flag signal is added to the node after branching. Flag signal confirmation 1540 confirms whether all added flag signals are included in the acquired integrated overview information, and accumulates the acquired integrated overview information until all flag signals can be confirmed. If all added flag signals have been validated, flag signal validation 1540 outputs the accumulated integration summary information to duplicate integration 1242 .

<Example of processing procedure of generation unit 121>
FIG. 16 is a flowchart of detailed generation processing procedures by the generation unit 121 of the design execution system 100 according to the fifth embodiment. FIG. 16 shows an example of generating configuration information using ID assignment, output number setting, duplicate integration, flag confirmation, etc. in the generation unit 121 of the design execution system 100 according to the fifth embodiment.

Only the differences from the fourth embodiment will be described in detail regarding the processing procedure for generating configuration information in the generation unit 121 of the design execution system 100 . The detailed processing procedure for generating configuration information differs from that of the fourth embodiment in that steps S1600, S1601, and S1602 are added to the flowchart of FIG.

At step S1600, the generation unit 121 determines whether the output of the node to be processed is branched (step S1600). If the output of the node is branched (step S1600: Yes), the process proceeds to step S1601, and if the output of the node is not branched (step S1600: No), the process proceeds to step S1405.

In step S1601, the generation unit 121 causes the node to output a flag signal to the subsequent node, and the subsequent node sets the acquired flag signal to the function overview information, and proceeds to the process of step S1405.

In step S1602, the generation unit 121 checks in flag signal confirmation 1440 whether all the added flag signals are included in the acquired integrated summary information. If all the added flag signals have been confirmed (step S1602: Yes), the generation unit 121 outputs the accumulated integrated summary information to the overlapping integration 1242, and proceeds to step S1408. If all the added flag signals cannot be confirmed (step S1602: No), generating section 121 accumulates the acquired integrated summary information, and proceeds to step S1400.

As described above, according to the fifth embodiment, after the nodes branched for parallelization are aggregated into one node once and the parallelization is aggregated, parallel processing or multi-process processing in the asynchronous system is re-parallelized. It is possible to obtain the design execution system 100 capable of executing synchronous processing.

Example 6 will be described with a focus on the differences from Examples 1 to 5. The same reference numerals are assigned to the points common to the first to fifth embodiments, and the description thereof is omitted. Embodiment 6 provides a system that enables execution of prioritized process processing in a system that includes parallel processing and multi-process processing in the generated configuration information 320 .

<Example of configuration information generation for system configuration including parallel processing>
FIG. 17 is an explanatory diagram of a detailed processing example of configuration information generation in the generation unit 121 of the design execution system 100 according to the sixth embodiment. Embodiment 6 differs from Embodiments 4 and 5 in that the function overview information includes two-dimensional coordinate information 1730 to 1740 in the workspace 240 where each node is arranged, and priority confirmation is performed in the execution node 1706. 1750 was added.

The two-dimensional coordinate information 1730 to 1740 are the two-dimensional coordinate positions in the workspace 240 where the nodes are arranged when assigning IDs to each node, and are added to the function overview information.

The priority confirmation 1750 confirms the position of each node in the acquired synchronous duplication/integration summary information group 1301, and preferentially processes the paths higher on the vertical axis of the two-dimensional coordinates in the work space 240. The process priority is added so as to be executed, and output to the configuration information generation 1244 as prioritized synchronous duplication integration summary information.

Specifically, for example, in the branch following camera input 1200, the node of reasoning A 1201 is located at the top in terms of arrangement, so processing with a higher priority is executed. This priority is also inherited by subsequent nodes. In other words, camera input 1200, reasoning A 1201, and result drawing 1205 are preferentially processed by other paths.

FIG. 18 is an explanatory diagram showing a specific example 1 of the processes of duplication integration 1242 and node connection arrangement 1243 according to the sixth embodiment. FIG. 18 shows a change in the integrated outline information from the integrated outline information group 1800 acquired by the overlap integration 1242 and the node connection arrangement 1243 in the execution node 1206 to the overlap composite outline information group 1801 .

For example, since the same ID "AAA" and the same two-dimensional coordinate information 1730-1732 are given to the function outline information 1220-1222 of the camera input 1200, the overlapping integration 1242 is determined to be the same node. , and output as overlapping and integrated summary information 1811 .

In addition, since the same ID "CCC" and the same two-dimensional coordinate information 1734, 1735 are assigned to the function overview information 1224 of the inference B1202 and the function overview information 1225 of the inference B1202, the overlapping integration 1242 is performed with the same node. It judges and integrates them, and outputs them as overlapping integrated overview information 1812 .

In addition, since the same ID "EEE" and the same two-dimensional coordinate information 1739, 1740 are assigned to the function outline information 1227 and 1228 of the result drawing 1205, the overlap integration 1242 determines that they are the same node and integrates them. Output as integrated overview information 1313 . The function outline information 1230 is also given the ID “EEE”, but the two-dimensional coordinate information 1738 is different from the two-dimensional coordinate information 1739 and 1740, so the function outline information 1230 is not integrated with the overlapping integrated outline information 1313 .

FIG. 19 is an explanatory diagram showing a specific example 2 of the processes of duplication integration 1242 and node connection arrangement 1243 according to the sixth embodiment. FIG. 19 shows a change in the integrated summary information from the overlapping composite summary information group 1801 to the synchronous duplicate integrated summary information group 1802 .

The node connection arrangement 1243 provides node connection confirmation and synchronization processing information 1920, and the priority confirmation 1750 includes duplicate integration overview information 1811 (camera input 1200), function overview information 1223 (inference A 1201), and function overview information 1226 ( A priority 1930 is given to the combination of result drawing 1205 ), that is, to the integrated summary information 1231 . The integrated outline information 1231 for which the priority 1930 is set is preferentially executed over the integrated outline information 1232 and 1233 . In this way, a prioritized synchronous duplication integration summary information group 1902 is generated. As a result, for example, it is possible to suppress the processing time and execute high-priority processing.

<Example of processing procedure of generation unit 121>
FIG. 20 is a flowchart of detailed generation processing procedures by the generation unit 121 of the design execution system 100 according to the sixth embodiment. FIG. 20 shows an example of generating configuration information using ID allocation, output number setting, duplicate integration, flag confirmation, etc. in the generation unit 121 of the design execution system 100 according to the sixth embodiment.

Regarding the detailed processing procedure for generating the configuration information in the generation unit 121 of the design execution system 100, only differences from the fourth and fifth embodiments will be described. The detailed processing procedure for generating configuration information differs from that of the fourth embodiment in that steps S2000 and S2001 are added.

In step S2000, the generation unit 121 acquires the two-dimensional coordinates of the position where the node is arranged in the work space 240, and outputs them to step S1404.

In step S2001, the generation unit 121 confirms the position of each node in the priority confirmation 1650, and performs processing so that the processing of the path that is higher on the vertical axis on the two-dimensional coordinates is preferentially executed. is added to generate prioritized synchronous duplication and integration overview information, and the process proceeds to step S1410.

As described above, according to the sixth embodiment, by assigning processing priority to nodes branched for parallelization from the node arrangement position, even in a processor with limited calculation speed and processing resources, a specific The design execution system 100 capable of suppressing the processing time for the processing of (1) can be obtained.

As described above, according to the design execution apparatus according to the first to sixth embodiments, the system to be designed can be designed using a graphical user interface, and the system can be executed even from the design base. Therefore, the system can be executed by the execution base alone at the time of execution after system design.

Also, the design execution devices according to the first to sixth embodiments described above can be configured as shown in (1) to (11) below.

(1) The design support device (client 110 ) has a design section 120 , a generation section 121 and an execution section 123 . The design unit 120 has, for each function, a node containing outline information defining information specifying the process without the entity of the process that realizes the function. Execution order data (system information) that defines the execution order of a specific function group related to the design object is generated by connecting a specific node group related to the object. In accordance with the execution order data generated by the design unit 120, the generation unit 121 inherits the outline information of the preceding node in the specific node group to the succeeding node connected to the succeeding node of the preceding node, thereby obtaining the specified node. The design object configuration information 320 is generated in which the outline information used in each of the function groups is defined according to the execution order. The execution unit 123 uses the configuration information 320 generated by the generation unit 121 to materialize processes for realizing the specific function group, and executes the materialized processes in the execution order.

As a result, it is possible to improve the convenience of designing and executing the design target. Specifically, for example, the design unit 120 and the generation unit 121 handle a group of nodes that do not have an entity of processing that implements a function, so only outline information is inherited between nodes. In other words, inter-node communication regarding the results of processing between nodes does not occur. Moreover, it is not necessary to pass data of processing results between the design unit 120, the generation unit 121, the configuration information 320 and the holding unit 122 that execute the processing for realizing the function of each node, and the execution base 432. FIG. Therefore, efficiency of design can be improved, and convenience for the designer is improved. In addition, since the execution unit 123 uses the configuration information 320 to materialize (instantiate) only the necessary functions, it is possible to improve the efficiency of the execution of the design target, thereby improving the convenience for the designer.

(2) In the design execution device of (1) above, the design unit 120 displays the nodes for each of the functions, and connects the plurality of nodes by an external operation (for example, a designer operating the input device 204). is possible, and the execution order data is generated by connecting the specific node group.

As a result, the design target can be designed using a graphical user interface, improving convenience for designers.

(3) In the design execution device of (2) above, the design section 120 displays detailed setting screens 246 to 248 as the outline information of the node by the external operation.

As a result, it is possible to visually recognize which function has which item and what the setting value is, which improves convenience for designers. In addition, by allowing the setting values to be changed by user operation, the convenience for the designer is further improved.

(4) In the design execution device of (2) above, the design section 120 causes the browser 400 of an external device that can communicate with the design execution device to display the nodes for each of the functions, and the external device from the external device. A plurality of nodes can be connected by an operation, and the execution order data is generated by connecting the specific node group.

This makes it possible to design and execute the design target remotely from an external device. Therefore, the convenience for the designer is improved.

(5) In the design execution device of (1) above, the generation unit 121 refers to the execution order data, and if there is the preceding node for each node of the specific node group, Integrated outline information is generated by integrating the outline information and the outline information of the node, and if the integrated outline information of the preceding node is generated by the preceding node, the integrated outline information of the preceding node and the outline information of the node , and outputs the integrated outline information generated at the final stage node as the configuration information 320 .

As a result, just by connecting the nodes by the designer, at each node, all the overview information from the top node to the relevant node is integrated and passed to the subsequent node. Therefore, when a designer simply connects nodes, a group of processes for realizing a group of specific functions are executed according to the order of execution. Therefore, the convenience for the designer is improved.

(6) In the design execution device of (5) above, the generating unit 121 adds a first identification that uniquely identifies the outline information within the integrated outline information to each of a plurality of pieces of outline information within the integrated outline information. If the information (ID) is not assigned, the first identification information (ID) is assigned (step S801).

As a result, when the designer simply connects the nodes, the first identification information is set in the outline information from the top node to the node in accordance with the execution order in each node. Processes that implement a particular set of functions are executed in sequence. Therefore, the convenience for the designer is improved.

(7) In the design execution device of (6) above, the generating unit 121 generates a plurality of preceding nodes for each node of the specific node group, and the plurality of preceding nodes have the same first node. If there is different summary information to which one identification information is assigned, second identification information (sub-ID) that distinguishes the different summary information is assigned to the first identification information (step S804).

As a result, the overlap of the first identification information is eliminated without the designer being conscious of the overlap of the first identification information simply by connecting the nodes, and parallel processing and multi-process processing can be realized. In addition, it is possible to prevent malfunctions in the execution unit 123 in advance. Therefore, the convenience for the designer is improved.

(8) In the design execution device of (1) above, the generating unit 121 generates a plurality of preceding nodes for each node of the specific node group, and the plurality of preceding nodes have the same outline information. exists, the duplication of the same outline information is eliminated and inherited by the succeeding node (step S803).

As a result, the designer can avoid redundant execution of the same process in the execution unit 123 by simply connecting the nodes without being aware of the duplication of the outline information. Therefore, efficiency of design can be improved, and convenience for the designer is improved.

(9) The design execution device of (1) above has a holding unit 122 that holds a function group 420 that is a set of the functions and the configuration information 320, and the execution unit 123 stores the above information in the holding unit 122. When the function group 420 and the configuration information 320 are held, the specific function group is selected from the function group 420 held in the holding unit 122, and the specific function group is selected using the configuration information 320. is materialized, and the materialized processes are executed in the execution order.

As a result, if the function group 420 and the configuration information 320 are held, the system can be executed by the execution unit 123 alone at the time of execution after system design without executing the design unit 120 and the generation unit 121. Therefore, efficiency of design can be improved, and convenience for the designer is improved. For example, during execution by the execution unit 123, there is no need to activate Nod-RED, which constitutes the design unit 120, and press the start button. Execution by the execution unit 123 is started only by turning on the power supply of the design execution device. Alternatively, the executing unit 123 may be started by starting from a specific program or by using a start signal as a signal.

(10) In the design execution device of (9) above, the holding unit 122 holds a containerized function group 1120 obtained by containerizing the function group 420, and the execution unit 123 is held in the holding unit 122. The specific containerized function group is selected from the containerized function group 1120, the specific containerized function group is materialized using the configuration information, and the materialized processes are executed in the execution order.

As a result, version management of the function group 420 becomes unnecessary, and malfunctions during execution of the execution unit 123 can be avoided. Such containerization facilitates maintenance of the function group 420 and improves convenience for the designer.

(11) In the design execution apparatus of (1) above, based on the process of realizing the processing device 203 that executes a specific process and the function of controlling the processing device 203 materialized by the execution unit 123, and a control unit 124 that controls the processing device 203 .

As a result, it is possible to implement the system designed while cooperating with the processing device 203 by materializing the functions described in the configuration information 320 .

(12) In the design execution device of (5) above, the generation unit 121 identifies the position of the node based on identification information (ID) that uniquely identifies the node and the number of the subsequent nodes to be connected. and the number of outputs indicating the number of execution paths of the execution order data to be executed, are stored in the summary information, and synchronization processing information is added to start processing when the output from the preceding node to the node is completed. By doing so, the configuration information 320 is generated.

As a result, synchronous processing (calling execution nodes) can be executed for parallel processing and multi-process processing in an asynchronous system.

(13) In the design execution device of (12) above, the generation unit 121 determines whether or not the integration overview information for the number of outputs has been acquired at the node of the final stage based on the number of outputs. , when acquired, the configuration information 320 is generated by adding the synchronization processing information.

This eliminates the need for a synchronous node to consolidate execution node calls from multiple execution paths into one call, and prevents human errors such as connection setting errors due to manual operation by the user.

(14) In the design execution device of (12) above, the generating unit 121 integrates a plurality of pieces of the outline information having the same identification information among the integrated outline information corresponding to the number of outputs to generate integrated outline information. The configuration information 320 is generated by generating and adding the synchronization processing information to the integrated outline information.

As a result, it is possible to reduce the amount of data in the configuration information 320 and simplify the execution order data.

(15) In the design execution device of (13) above, when the node branches to a plurality of the later nodes, the generation unit 121 acquires a flag signal regarding the number of branches for each of the latter nodes, and in the final stage node, based on the flag signal and the number of outputs, it is determined whether or not the integrated summary information for the number of outputs has been acquired. The configuration information 320 is generated by adding the synchronization processing information.

As a result, the nodes branched for parallelization are once aggregated into one node, and after the parallelization is aggregated, even in the asynchronous system that is parallelized again, synchronous processing (calling the execution node) can be executed. can.

(16) In the design execution device of (12) above, the generation unit 121 acquires coordinate information indicating the coordinate positions where each node of the specific node group is arranged in the workspace 240 by the external operation, It is held in the summary information of the node.

As a result, it is possible to confirm which node is placed at which position on the workspace 240.

(17) In the design execution device of (16) above, the generation unit 121 sets the priority of a plurality of execution paths included in the execution order data based on the coordinate information in the final stage node. Thus, the configuration information 320 is generated.

As a result, a priority can be set for each execution path, and processes related to execution paths can be executed according to the priority.

The present invention is not limited to the embodiments described above, and includes various modifications and equivalent configurations within the scope of the attached claims. For example, the above-described embodiments have been described in detail in order to facilitate understanding of the present invention, and the present invention is not necessarily limited to those having all the described configurations. Also, part of the configuration of one embodiment may be replaced with the configuration of another embodiment. Moreover, the configuration of another embodiment may be added to the configuration of one embodiment. Moreover, other configurations may be added, deleted, or replaced with respect to a part of the configuration of each embodiment.

Further, each configuration, function, processing unit, processing means, etc. described above may be realized by hardware, for example, by designing an integrated circuit, for example, by designing a part or all of them, and the processor 201 performs each function. It may be realized by software by interpreting and executing a program to be realized.

Information such as programs, tables, files, etc. that realize each function is recorded on storage devices such as memory, hard disk, SSD (Solid State Drive), or IC (Integrated Circuit) card, SD card, DVD (Digital Versatile Disc) Can be stored on media.

In addition, the control lines and information lines indicate those that are considered necessary for explanation, and do not necessarily indicate all the control lines and information lines necessary for implementation. In fact, it can be considered that almost all configurations are interconnected.

Claims (21)

1. Each function has a node containing outline information specifying information specifying the process without the entity of the process that realizes the function, and connection of a plurality of the nodes is possible by external operation, and connection of a specific node group a design unit that generates execution order data that defines the execution order of a specific function group by
   According to the execution order data generated by the design unit, each of the specific function groups inherits the summary information of the preceding node of the specific node group to the succeeding node connected after the preceding node. a generation unit that generates configuration information defining the summary information to be used according to the execution order;
   an execution unit that materializes processes for realizing the specific function group using the configuration information generated by the generation unit and executes the materialized processes in the execution order;
   A design execution device comprising:
2. The design execution device according to claim 1,
   The design unit displays the nodes for each of the functions, is capable of connecting a plurality of the nodes by an external operation, and generates the execution order data by connecting the specific node group.
   A design execution device characterized by:
3. The design execution device according to claim 2,
   The design unit displays the summary information of the node by the external operation.
   A design execution device characterized by:
4. The design execution device according to claim 2,
   The design unit causes an external device that can communicate with the design execution device to display the nodes for each of the functions, and allows connection of a plurality of the nodes by the external operation from the external device. generating the execution order data by connecting groups;
   A design execution device characterized by:
5. The design execution device according to claim 1,
   The generating unit refers to the execution order data, and for each node of the specific node group, if there is the preceding node, an integrated summary obtained by integrating the summary information of the preceding node and the summary information of the node. information is generated, and if the integrated summary information of the preceding node is generated by the preceding node, the integrated summary information is generated by integrating the integrated summary information of the preceding node and the summary information of the node, and the final step of outputting the integrated overview information generated by the node as the configuration information;
   A design execution device characterized by:
6. The design execution device according to claim 5,
   The generating unit assigns the first identification information to each of a plurality of pieces of overview information in the integrated overview information if the first identification information that uniquely identifies the overview information in the integrated overview information is not assigned. ,
   A design execution device characterized by:
7. The design execution device according to claim 6,
   If each node of the specific node group has a plurality of preceding nodes and different summary information to which the same first identification information is assigned to the plurality of preceding nodes, , assigning to the first identification a second identification that distinguishes between the different summary information;
   A design execution device characterized by:
8. The design execution device according to claim 1,
   If each node of the specific node group has a plurality of preceding nodes and the same summary information exists in the plurality of preceding nodes, the generation unit eliminates duplication of the same summary information. and inherit to the latter node,
   A design execution device characterized by:
9. The design execution device according to claim 1,
   a holding unit that holds a function group, which is a set of the functions, and the configuration information;
   When the holding unit holds the function group and the configuration information, the execution unit selects the specific function group from among the function groups held in the holding unit, and uses the configuration information to perform materializing processes for realizing the specific function group and executing the materialized processes in the execution order;
   A design execution device characterized by:
10. The design execution device according to claim 9,
    The holding unit holds a containerized function group obtained by containerizing the function group,
    The execution unit selects the specific containerized function group from the containerized function groups held in the holding unit, materializes the specific containerized function group using the configuration information, and materializes the containerized function group. executing the processes in the order of execution;
    A design execution device characterized by:
11. The design execution device according to claim 1,
    a processing device that performs a specific process;
    a control unit that controls the processing device based on a process that realizes a function of controlling the processing device materialized by the execution unit;
    A design execution device comprising:
12. The design execution device according to claim 5,
    The generation unit includes identification information that uniquely identifies the node, an output number that indicates the number of execution paths of the execution order data that is identified by the position of the node based on the number of the subsequent nodes that are connected, and is obtained, held in the summary information, and synchronous processing information for starting processing when output from the preceding node to the node is completed, thereby generating the configuration information.
    A design execution device characterized by:
13. 13. The design execution device of claim 12,
    The generation unit determines whether or not the integrated summary information for the number of outputs has been acquired at the node in the final stage based on the number of outputs, and if acquired, adds the synchronization processing information , generating said configuration information;
    A design execution device characterized by:
14. 13. The design execution device of claim 12,
    The generating unit generates integrated outline information by integrating a plurality of pieces of outline information having the same identification information among the integrated outline information corresponding to the number of outputs, and adds the synchronization processing information to the integrated outline information. generating the configuration information by
    A design execution device characterized by:
15. 14. The design execution device of claim 13,
    The generation unit acquires a flag signal regarding the number of branches for each of the post-nodes when the node branches to the plurality of post-stage nodes, holds the flag signal in the summary information of the post-stage node, and determining whether or not the integrated summary information for the number of outputs has been acquired based on the flag signal and the number of outputs; generate,
    A design execution device characterized by:
16. 13. The design execution device of claim 12,
    The generating unit acquires coordinate information indicating a coordinate position where each node of the specific node group is arranged in the work area by the external operation, and holds it in the summary information of the node.
    A design execution device characterized by:
17. 17. The design execution device of claim 16,
    The generation unit generates the configuration information by setting priorities of a plurality of execution paths included in the execution order data based on the coordinate information at the final node.
    A design execution device characterized by:
18. A design execution system comprising a server holding a function group and a client capable of communicating with the server,
    Each function has a node containing outline information specifying information specifying the process without the entity of the process that realizes the function, and connection of a plurality of the nodes is possible by external operation, and connection of a specific node group a design unit that generates execution order data that defines the execution order of a specific function group by
    According to the execution order data generated by the design unit, each of the specific function groups inherits the summary information of the preceding node of the specific node group to the succeeding node connected after the preceding node. a generation unit that generates configuration information defining the summary information to be used according to the execution order;
    an execution unit that materializes processes for realizing the specific function group by using the configuration information generated by the generation unit and executes the materialized processes in the execution order;
    A design execution system comprising:
19. 19. The design execution system of claim 18, comprising:
    Said client
    the design unit, the generation unit, and the execution unit;
    A design execution system comprising:
20. 19. The design execution system of claim 18, comprising:
    The server is
    having the design unit and the generation unit;
    Said client
    having the execution unit;
    A design execution system characterized by:
21. A computer having a processor that executes a program and a storage device that stores the program,
    Each function has a node containing outline information specifying information specifying the process without the entity of the process that realizes the function, and connection of a plurality of the nodes is possible by external operation, and connection of a specific node group A design process for generating execution order data that defines the execution order of a specific function group by
    According to the execution order data generated by the design process, each of the specific function groups inherits the outline information of the preceding node of the specific node group to the succeeding node connected to the succeeding of the preceding node. a generating process for generating configuration information defining the summary information to be used according to the execution order;
    an execution process of materializing processes for realizing the specific function group using the configuration information generated by the generation process, and executing the materialized processes in the execution order;
    A design execution method characterized by executing

Images