KR20200013543A - Apparatus and method for constructing virtual factory - Google Patents

Abstract

The present invention relates to a device for constructing a virtual factory. The device comprises: a clustering module which performs clustering by classifying a virtual space into regions using location information of a plurality of objects received from an indoor positioning device; an estimation module estimating what process is performed in the clustered region; and a rule tracking module tracking rules with regard to the estimated process using process reference information related to a product production process received from a production system.

Description

Device and method for building virtual factory {APPARATUS AND METHOD FOR CONSTRUCTING VIRTUAL FACTORY}

The present invention relates to an apparatus and method for building a virtual factory, and more particularly, to an apparatus and method for constructing a virtual factory using artificial intelligence based on indoor positioning data and process data collected in a real factory.

A virtual factory is a virtual world modeled within a process simulator that can describe the manufacturing process of a real factory. Virtual factories are technologies that can reduce the cost of preparing alternatives to various production plans, and can test and evaluate alternatives of production plans in advance, and are increasingly important to improve manufacturing productivity in recent years. However, the construction of a virtual factory requires a large number of professional personnel and a problem that requires a high cost of building a virtual factory. Therefore, a low cost virtual factory building technology is required.

Published Patent Publication 10-2009-0123051

The problem to be solved by the present invention is to provide a virtual factory building apparatus and method that does not require a large number of skilled personnel, the cost of building.

According to an aspect of the present invention, there is provided a virtual factory building device comprising: a clustering module for performing clustering by classifying a virtual space into an area by using location information of a plurality of objects received from an indoor positioning device. An estimation module for estimating what processes are performed in the clustered region, and a rule tracking module for tracking rules for the estimated processes using process reference information related to the product production process received from the manufacturing system.

It may further include an artificial intelligence module for performing artificial intelligence learning using the virtual factory building data accumulated while building the virtual factory.

Each of the clustered regions may have the same or similar pattern in time and space in the virtual space.

The estimating module may estimate a process of the clustered area by using feature information about the clustered area.

The rule tracking module may verify and correct the estimated process using the process reference information.

According to another embodiment of the present invention, a virtual factory building system is configured to communicate with a tag attached to an object to generate location information for identifying the location of the object, and to classify clustering by classifying a virtual space into an area. And a virtual factory building apparatus for estimating what processes are performed in the clustered region and tracking rules for the estimated processes using process reference information related to a product production process received from a manufacturing system.

According to another embodiment of the present invention, a virtual factory building method for building a virtual factory using location information of a plurality of targets received from an indoor positioning device includes: classifying a virtual space into an area and performing clustering. Estimating what processes are performed in the region, and tracking rules for the estimated processes using process reference information related to the product production process received from the manufacturing system.

As described above, according to the apparatus and method for constructing a virtual factory according to an embodiment of the present invention, a virtual factory corresponding to a real factory can be constructed using only indoor positioning data collected from a real factory based on indoor positioning technology and artificial intelligence technology. As virtual factory building data accumulates, artificial intelligence can take the place of artificial intelligence rather than human experts. Therefore, the cost and time required to build a virtual factory can be reduced, and it can be applied to various job groups by industry, factory, and process.

1 is a schematic diagram illustrating a virtual factory building system according to an embodiment of the present invention.
FIG. 2 is a schematic view for explaining the indoor positioning device shown in FIG. 1.
3 is a block diagram of the virtual factory building device shown in FIG. 1.
Figure 4 is a schematic diagram for explaining the clustering step of building a virtual factory according to an embodiment of the present invention.
5 is a schematic diagram for explaining an estimating step during virtual plant construction according to an embodiment of the present invention.
6 is a schematic diagram for explaining a rule inversion step of building a virtual factory according to an embodiment of the present invention.
7 is a flowchart illustrating a virtual factory building method according to an embodiment of the present invention.

The terms or words used in this specification and claims are not to be construed as limiting in their usual or dictionary meanings, and the inventors may appropriately define the concepts of terms in order to best describe their invention. It should be interpreted as meanings and concepts corresponding to the technical idea of the present invention based on the principle that the present invention.

Therefore, the embodiments described in the specification and the configuration shown in the drawings are only the most preferred embodiments of the present invention and do not represent all of the technical idea of the present invention, various equivalents that may be substituted for them at the time of the present application It should be understood that there may be variations and examples.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention.

1 is a schematic diagram illustrating a virtual factory building system according to an embodiment of the present invention, FIG. 2 is a schematic diagram illustrating an indoor positioning device, and FIG. 3 is a block diagram of a virtual factory building device according to an embodiment of the present invention. It is also. 1 to 3, the virtual factory building system according to the embodiment of the present invention includes a virtual factory building device 100 and an indoor positioning device 200, and is installed in the real factory 10.

The indoor positioning device 200 includes a plurality of anchors 210 and a plurality of tags 220. The plurality of anchors 210 are installed on the ceiling or wall of the interior of the reality factory 10, and the plurality of tags 220 are attached to the plurality of objects 12 to determine the position in the reality factory 10, respectively. Here, the object 12 refers to a product produced in the actual factory 12, but does not mean only the finished product, but may be a material or a semi-finished product in the production process, and if the shape can be attached to the tag 220, It doesn't matter. These objects 12 are made in the same place as the work bench 11 provided in the real factory 10 and moved through means such as the conveyor 13.

The plurality of anchors 210 performs wireless communication with the plurality of tags 220 to continuously measure the position of each tag 220 in real time. Therefore, the indoor positioning device 200 can know the present location as well as the past location of each object 12, and the location of the target object 12 is determined from the time when each object 12 is started for the first time until the work is finished. I can figure it out.

If the location of the object 12 can be determined, the indoor positioning device 200 may use any type of communication method such as WiFi, Bluetooth, Beacon, Zigbee, or magnetic field, and any type of AOA, ROA, Finger Print, TDOA, TOA, etc. It is also possible to use a positioning algorithm.

The virtual factory building device 100 receives the identification information and the location information of each object 12 from the indoor positioning device 200, and uses the spaces between the objects 12, the moving speed and the moving direction of the objects 12. For example, the time during which each object 12 stays at a specific position can be calculated. The virtual factory building apparatus 100 builds a virtual factory using the calculated information.

Referring to FIG. 3, the virtual factory building device 100 includes a clustering module 110, an estimation module 120, a rule inversion module 130, and an artificial intelligence module 140, each of which is clustered ( The virtual factory is built by performing the clustering step, the identifying step, the rule tracing step, and the AI learning step.

Figure 4 is a schematic diagram for explaining the clustering step of the virtual plant construction according to an embodiment of the present invention, Figure 5 is a schematic diagram for explaining the estimation step during the construction of a virtual factory according to an embodiment of the present invention, Figure 6 It is a schematic diagram for explaining a rule inversion step during the construction of a virtual factory according to an embodiment of the present invention.

First, the virtual factory building apparatus 100 assumes a virtual space in which the virtual factory is to be constructed using all the location information of the entire target 12. That is, it creates a virtual space that can contain all the locations where product 12 can exist.

Thereafter, the clustering module 110 performs clustering by classifying spaces in which virtual objects have the same or similar characteristics or patterns in time and space in the virtual space with reference to the location information of each object 12. For example, referring to FIG. 4, the clustering module 110 divides the virtual space into regions A, B, C, D, and E to perform clustering. Marked with dots in FIG. 4 represents the object 12 located in the virtual space at any point in time.

Area A has a feature that the object 12 located there stays for a certain time or more, a certain number of objects 12 are gathered at the same time zone, and the moving distance of the object 12 is very short and leaves only one direction. Is characterized by each object 12 moving in one direction with a constant speed, region C is characterized by each object 12 staying there for a certain time and moving in either of two moving directions, The area D moves at a constant speed but is bent at a certain position by 90 degrees, and the area E has a certain number of objects 12 moved at a time after the object 12 stays there for a predetermined time, and a small number of objects 12 ) Is left in a separate space.

Clustering as shown in FIG. 4 is exemplary, and in a real factory, since the temporal and spatial characteristics of the object 12 may appear more diverse, more various criteria having the same or similar characteristics may be presented for clustering.

When the clustering is completed, the estimation module 120 estimates what kind of process the clustered area is. That is, it estimates what process is performed in the clustered area by using the characteristic information of the clustered area.

Referring to FIG. 5, as in the previous example of FIG. 4, in the case of region A, the target 12 located there stays for a predetermined time or more, and a predetermined number or more of the targets 12 are gathered at the same time zone. Because it is very short and has only one way out of it, it can be assumed that the process in area A is a machining process. In addition, in the case of the region B, each object 12 has a characteristic of moving at a constant speed in one direction, so that the process performed in the region B can be estimated as a linear movement process. Similarly, in the case of the region C, since each object 12 has a characteristic of staying there for a predetermined time and moving in one of two moving directions, it can be estimated that the process performed in the region C is a classification process. In the case of the region D, since the object 12 moves at a constant speed but is bent at 90 degrees from a specific position, the process performed in the region D may be assumed to be a "letter movement process". In the case of the region E, since the object 12 stays there for a predetermined time or more, a certain number or more of the objects 12 are moved at once and a few objects 12 are left in separate spaces. It can be estimated that it is an inspection process.

The working processes shown in FIG. 5 are exemplary and more various processes may be included as the estimating process in the estimating step.

In this way, the estimation module 120 may simply estimate what process is performed for any clustered area, but the clustering process is defined in advance for various work processes performed on the object 12 in the actual factory. The estimating step may be performed by matching what work is done on the area.

Although the clustering step and the estimating step are separated, the clustering step is performed, and then the estimating step is performed. However, the clustering step and the estimating step are not necessarily required.

As such, when the virtual factory building device 100 performs clustering and estimation using location information from the indoor positioning device 200, a skeleton of a virtual factory, that is, a map of a virtual factory is made.

The virtual factory building apparatus 100 may receive process reference information of the real factory 10 from a manufacturing system (not shown) in advance with respect to the virtual factory building. The process reference information includes various information related to the production of the object 12 as well as information about the object 12 itself. It may include values measured at the sensor, equipment values, processing time, travel time, standby time, and the like.

The rule inversion module 130 constructs a more accurate virtual factory by inverting the rules in the real factory 10 by linking process reference information with respect to the previously clustered and estimated processes. That is, the rule inversion module 130 may verify and correct the previously estimated process using the process reference information, and add the flesh of the virtual factory by adding a rule or the like to the estimated process.

For example, referring to FIG. 6, the product A is disposed in the area E on the left side, the product B is disposed in the area E on the right side, and the product A is the quality inspection standard according to the process standard information. If a, b is to be satisfied, and product B is to meet the quality inspection criteria c and d, this content is added to the quality inspection process of the virtual factory by the rule inversion module 130. In addition, in the classification process of the area C, when the classification criteria for the A product and the B product are A specific and B characteristics, these contents are added to the virtual factory classification process. In the linear movement process of the area B, it is possible to verify whether the process is correct by comparing the time during which the object 12 moves the area B with the movement time shown in the process reference information. 12) It is possible to verify whether the process is correct by comparing the time remaining in this area A with the processing time of the process reference information.

As such, the rule inversion module 130 may complete the virtual factory by verifying and adding various rules to the virtual factory on which the skeleton is completed by using the process reference information. On the other hand, the rule inversion module 130 has been described as following the rule upside down to what process has been passed to the finished product for the finished product, but it is not necessary to do so, it is not necessary to do so sequentially from the time when the initial material is input You can also follow the rules.

The artificial intelligence module 140 performs artificial intelligence learning to more accurately build a virtual factory and exclude the role of a person. Initially, even if a virtual factory is constructed by receiving the location information of the object 12 from the real factory 10, a considerable portion may not fit well. In this case, it is possible to go through a process of correcting the built virtual factory, such as teaching about the unsuitable part of the human expert or the real factory 10 concerned. Through this process, the artificial intelligence module 140 learns an accurate process or rule corresponding to the location information of the object 12.

In the early stages of building a virtual factory, a person could build a virtual factory for artificial intelligence learning. That is, a human expert reports that the virtual factory building device 200 displays the location of the object 12 in a virtual space on a display device such as a monitor using the location information of the object 12 as shown in FIG. 4. Inversion can be performed to build a virtual factory. Since a person has intuitive power, the flow of the product 12 can be clustered, the process can be estimated by processing, linear movement, sorting, etc., and the average time required for the processing process, For the linear movement process, a regular inversion, such as how much the speed is on average and in which classification the classification process is performed, may be performed.

As shown in the following [Table 1], illustrating the role of the artificial intelligence module 140 according to the number of virtual factory building, the first part of the expert intervention, but the data is accumulated as the virtual factory building cycle is repeated. Over time, the AI module 140 may replace the role of the expert. In addition, by collecting virtual plant construction data by industry, by plant, and by process, the cycle repeats and the role of artificial intelligence grows over time.

TABLE 1

The artificial intelligence module 140 may use any algorithm using artificial neural network or machine learning other than this, including deep learning.

Next, a virtual factory building method according to another embodiment of the present invention will be described with reference to FIG. 7. 7 is a flowchart illustrating a virtual factory building method according to an embodiment of the present invention.

Since the virtual factory building method according to the embodiment of the present invention is executed by the virtual factory building device 100 described above, a detailed description thereof will be omitted and the operation of the virtual factory building device 100 described above will apply mutatis mutandis.

First, the virtual factory building device 100 receives the identification information and the location information of the object 12 from the indoor positioning device 200 and uses the spaces between the objects 12, the moving speed and the moving direction of the object 12, The amount of time each object 12 stays at a specific position is calculated.

Thereafter, the virtual factory building apparatus 100 performs clustering by classifying spaces in which virtual objects have the same or similar characteristics or patterns in time and space in the virtual space by referring to the location information of each object 12. (S710).

When the clustering is completed, the virtual factory building apparatus 100 estimates what kind of processes are targeted for the clustered area (S720). That is, it estimates what process is performed in the clustered area by using the characteristic information of the clustered area.

The virtual factory building apparatus 100 builds a more accurate virtual factory by inverting the rules in the real factory 10 by interworking process reference information with respect to the previously clustered and estimated processes (S730). That is, the virtual factory building apparatus 100 may verify and correct the previously estimated process by using the process reference information, and may add the flesh of the virtual factory by adding a rule or the like to the estimated process.

The virtual factory building apparatus 100 may perform the artificial intelligence learning to more accurately build the virtual factory and exclude the role of a person (S740). When human experts are involved, the data accumulated when the virtual factory is built can be used to learn precise processes or rules corresponding to the location information of the object 12.

The foregoing detailed description illustrates the present invention. In addition, the foregoing description merely shows and describes preferred embodiments of the present invention, and the present invention can be used in various other combinations, modifications, and environments. That is, changes or modifications may be made within the scope of the concept of the invention disclosed in this specification, the scope equivalent to the disclosed contents, and / or the skill or knowledge in the art. The above-described embodiments are intended to illustrate the best state in carrying out the present invention, the use of other inventions such as the present invention in other state known in the art, and the specific fields of application and uses of the present invention. Various changes are also possible. Thus, the detailed description of the invention is not intended to limit the invention to the disclosed embodiments. Also, the appended claims should be construed to include other embodiments.

10: reality factory, 12: destination,
100: virtual factory building device,
110: clustering module, 120: estimation module,
130: rule inversion module, 140: AI module,
200: indoor positioning device,
210: anchor, 220: tag

Claims (15)

Clustering module for performing the clustering by classifying the virtual space into the area by using the location information of the plurality of targets received from the indoor positioning device,
An estimation module for estimating what processes are performed in the clustered region, and
Rule tracking module for tracking rules for the estimated process using process reference information related to product production processes received from a manufacturing system
Virtual factory building device comprising a. In claim 1,
The virtual factory building device further comprising an artificial intelligence module for performing artificial intelligence learning using the virtual factory building data accumulated while building the virtual factory. In claim 1,
The clustered area is a virtual factory building device of the virtual space, each object is a space having the same or similar pattern in time and space. In claim 1,
And the estimating module estimates a process of the clustered area by using feature information about the clustered area. In claim 1,
The rule tracking module is a virtual factory building device for verifying and correcting the estimated process using the process reference information. An indoor positioning device that communicates with a tag attached to an object to generate location information to determine the location of the object; and
Classify virtual space into regions to perform clustering, estimate what processes are performed in the clustered regions, and track rules for the estimated processes using process reference information related to product production processes received from a manufacturing system. Virtual factory building device
Virtual factory building system comprising a. In claim 6,
The virtual factory building device is a virtual factory building system for performing artificial intelligence learning using the virtual factory building data accumulated while building a virtual factory. In claim 6,
The clustered area is a virtual factory building system in which each object in the virtual space is a space having the same or similar pattern in time and space. In claim 6,
And the virtual factory building apparatus estimates a process of the clustered area by using feature information on the clustered area. In claim 6,
And the virtual factory building device verifies and corrects the estimated process using the process reference information. A virtual factory construction method for establishing a virtual factory using location information of a plurality of objects received from an indoor positioning device,
Classifying the virtual space into regions and performing clustering;
Estimating what process is performed in the clustered region, and
Tracking rules for the estimated process using process reference information related to the product production process received from the manufacturing system
Virtual factory building method comprising a. In claim 11,
And performing artificial intelligence learning using the virtual factory building data accumulated while building the virtual factory. In claim 11,
The clustered region is a virtual factory building method in which each object in the virtual space is a space having the same or similar pattern in time and space. In claim 11,
And the estimating step includes estimating a process of the clustered area by using feature information on the clustered area. In claim 11,
The rule tracking step is a virtual factory building method for verifying and correcting the estimated process using the process reference information.

Images