JP7390205B2 - Process management method for ground improvement work using artificial intelligence (AI) and process management system for ground improvement work using artificial intelligence (AI) - Google Patents

Description

Application of Article 30, Paragraph 2 of the Patent Act (1) Publication date: February 25, 2019 Publication destination: ITbook Holdings Co., Ltd. website https://ssl4. eir-parts. net/doc/1447/tdnet/1679130/00. pdf Publisher ITbook Holdings Co., Ltd. (2) Publication date February 25, 2019 Publication destination TSE Mothers website https://www2. tse. or. jp/disc/14470/140120190225481422. pdf Publisher ITbook Holdings Co., Ltd.

The present invention relates to a method for managing the process of ground improvement work using artificial intelligence (AI), a system for implementing the method, and a jig used for the method.

In ground improvement work called the surface mixing method, a cement-based solidifying material is sprinkled on soft ground, followed by stirring, mixing, and compaction. Ground improvement has been carried out by pumping a solidification material slurry containing .

In this ground improvement work, normal ground improvement is carried out after performing the ground improvement process of mixing and stirring cement-based solidification material into the ground to be improved, or the ground improvement process of pumping the solidification material slurry and stirring and mixing it with soil. In order to perform a quality inspection to determine whether or not the ground improvement work has been carried out, test cores of the ground to be improved are collected by boring, and the quality of the collected test cores is tested to determine whether the ground improvement work is good or not. .

Perform the quality test using the following steps. Two to three days after the soil improvement process is completed, boring is carried out and test cores are collected, and the collected test cores are sent to a quality testing laboratory. Thereafter, first, 7 days after collection, a preliminary test of a uniaxial compression test of the test cores is conducted. As a result of the preliminary test, if the test result at the age of 7 days is equal to or higher than the design standard strength, the test core will be stored until 28 days have passed after collection, and after 28 days, it will be subjected to the uniaxial compression test. The main test will be conducted. If the test core shows good strength as a result of the main test, it is finally determined that the ground improvement work has been carried out successfully.

Here, when conducting the preliminary test or the main test, the test core is a core that is not suitable for the quality test (defective core), for example, a core that has cracks that are less than the length of the test specimen. , the core is mixed with soil clods, has no shape, etc., and uniaxial compression tests cannot be performed. If the test core is found to be such a defective core, it is necessary to perform boring again on the ground where the ground improvement work has been performed and re-collect the test core.

Then, the period from when the test core is first collected to when it is re-collected, that is, 7 days if the core is found to be defective in the preliminary test; Construction work following ground improvement work will not be possible for 28 days, leading to delays in the construction period.

For this reason, before sending a test core to a quality testing laboratory, it is necessary to determine whether the core is suitable for testing or is defective. Currently, after collecting test cores by boring, site managers and designers check the condition of the test cores visually or by sketching the cores to determine whether the cores are suitable for testing or are defective. It is determined whether If the sampled test core is determined to be suitable for testing, the test core is sent to a quality testing laboratory, and on the other hand, if it is determined to be a defective core, it is sent again. Perform boring to collect cores for testing.

However, until now, site managers and designers had to visually determine whether a core is suitable for testing or a defective core, so there was a lot of variation depending on the person making the decision, and whether a core was suitable for testing or not. Since it is difficult to accurately determine whether a core is defective or defective, there is a problem in that incorrect determination may be made. In addition, because of this, a large amount of experience is required for the discriminator, and it takes time to train skilled discriminators, so there is a problem that there is a shortage of discriminators who can make accurate discriminations. Ta.

Therefore, an object of the present invention is to enable, in process control of ground improvement work, to determine whether a test core collected by boring is a core suitable for testing or a defective core, without being limited by the experience of the discriminator. The purpose is to provide a method that can be used accurately.

The above-mentioned problem is solved by the following present invention.
That is, the present invention (1) is stored in the storage unit of a server in an artificial intelligence (AI), is collected within 168 hours after the end of the ground improvement work, and is collected at least 72 hours after the end of the ground improvement work. By training the image data group of normal cores suitable for quality testing and defective cores not suitable for quality testing as learning data, which were imaged after a learning step of storing a data group shown in the storage unit;
Imaging in which an image of the test core to be judged is taken from the ground where the ground improvement work has been performed, within 168 hours after the end of the ground improvement work, and is taken at least 72 hours after the end of the ground improvement work. process and
a transmission step of transmitting image data of the test core to be determined obtained through the imaging step to the server;
Artificial intelligence (AI) compares the image data of the test core to be determined, received by the reception unit of the server, with the data group indicating the defective core stored in the storage unit, and determines the target to be determined. an analysis step of determining whether or not the test core corresponds to a defective core;
a confirmation step of confirming the analysis results in the analysis step;
to have,
The present invention provides a process management method for ground improvement work using artificial intelligence (AI), which is characterized by:

In addition, the present invention ( 2 ) is a process management system for ground improvement work using artificial intelligence (AI) for carrying out the process control method for ground improvement work using artificial intelligence (AI) of (1),
At least an image of the test core to be determined, taken from the ground where the ground improvement work was performed, within 168 hours after the end of the ground improvement work, and taken at least 72 hours after the end of the ground improvement work. It has a transmitting device that transmits data to a server, and a server that connects to the transmitting device via the Internet,
The server is
a receiving unit that receives image data of the test core to be determined sent from the transmitting device;
A data group of images of normal cores suitable for quality testing and defective cores not suitable for quality testing, collected within 168 hours after the completion of soil improvement work and taken at least 72 hours after the completion of ground improvement work. A memory unit that stores information,
Normal cores and quality test that are stored in the storage unit of the server and are collected within 168 hours after the end of the ground improvement work and are imaged at least 72 hours after the end of the ground improvement work and are suitable for quality testing. learns a group of image data of defective cores that are not suitable as learning data, generates a group of data indicating defective cores, stores the generated data group indicating defective cores in the storage unit, and a receiving unit of the server; Compares the image data of the test core to be determined received with the data group indicating a defective core stored in the storage unit, and determines whether or not the test core to be determined corresponds to a defective core. Artificial intelligence (AI) that determines whether
to have
This provides a process management system for ground improvement work using artificial intelligence (AI), which is characterized by:

According to the present invention, in the process management of ground improvement work, it is possible to accurately determine whether a test core collected by boring is a core suitable for testing or a defective core, without depending on the experience of the judge. We can provide a method that can be used.

FIG. 1 is a schematic perspective view showing an imaging jig according to a first embodiment of the present invention. 2 is an enlarged view of the vicinity of the base portion 22 in FIG. 1. FIG. FIG. 2 is a diagram showing a state in which the shaft portion 25 of the imaging jig in FIG. 1 is contracted. It is a typical perspective view showing the imaging jig of the second form of the present invention. It is a figure which shows the tower part 55 in FIG. 5 is a cross-sectional view of the mounting portion 56 in FIG. 4. FIG. 5 is a top view of the mounting section 56 in FIG. 4. FIG. FIG. 5 is a cross-sectional view showing an example of an attachment form of a mounting portion 56. FIG. FIG. 5 is a cross-sectional view showing an example of an attachment form of a mounting portion 56. FIG. FIG. 5 is a cross-sectional view showing an example of an attachment form of a mounting portion 56. FIG. FIG. 5 is a cross-sectional view showing an example of an attachment form of a mounting portion 56. FIG. It is a typical perspective view showing the imaging jig of the second other form of the present invention. FIG. 2 is a schematic diagram showing an imaging box related to the testing core storage device of the present invention. FIG. 2 is a cross-sectional view of an example of the test core storage device of the present invention taken in the direction of the rotation axis. 15 is a perspective view showing the dish portion 64 in FIG. 14. FIG. FIG. 15 is a cross-sectional view of the test core storage device 61 in FIG. 14 taken in a direction perpendicular to the rotation axis. FIG. 15 is a cross-sectional view of the test core storage device 61 in FIG. 14 taken in a direction perpendicular to the rotation axis. FIG. 15 is a cross-sectional view of the test core storage device 61 in FIG. 14 taken in a direction perpendicular to the rotation axis. FIG. 15 is a cross-sectional view of the test core storage device 61 in FIG. 14 taken in a direction perpendicular to the rotation axis.

The process management method for the ground improvement process of the present invention is such that the information is collected using artificial intelligence (AI), is stored in the storage section of the server, is collected within 168 hours after the end of the ground improvement work, and is collected at least from the time of the end of the ground improvement work. A group of image data of a normal core suitable for a quality test and a defective core not suitable for a quality test, which were imaged after 72 hours, were trained as learning data to generate a data group indicating a defective core. a learning step of storing a data group indicating a defective core in the storage unit;
Imaging in which an image of the test core to be judged is taken from the ground where the ground improvement work has been performed, within 168 hours after the end of the ground improvement work, and is taken at least 72 hours after the end of the ground improvement work. process and
a transmission step of transmitting image data of the test core to be determined obtained through the imaging step to the server;
Artificial intelligence (AI) compares the image data of the test core to be determined, received by the reception unit of the server, with the data group indicating the defective core stored in the storage unit, and determines the target to be determined. an analysis step of determining whether or not the test core corresponds to a defective core;
a confirmation step of confirming the analysis results in the analysis step;
to have,
This is a process management method for ground improvement work that uses artificial intelligence (AI). Hereinafter, artificial intelligence (AI) will be abbreviated as AI.

The process management method for ground improvement work of the present invention includes a learning process. In the learning process, the AI is made to learn a group of image data of normal cores and defective cores stored in the storage unit of the server as learning data to generate a data group indicating a defective core, and to generate a data group indicating a defective core. This is a step of storing a data group indicating .

In the learning process, the image data group that the AI learns was collected within 168 hours after the completion of the ground improvement work by boring the target ground after the ground improvement process in order to conduct quality tests in the past. , is an image of the test core taken at least 72 hours after the end of the ground improvement work, and is an image of the preliminary test of the quality test 7 days after collection or the main test of the quality test 28 days after collection. Sometimes, a group of image data of test cores is stored in the storage unit of the server, categorized into normal cores that are determined to be suitable for quality testing, and defective cores that are determined to be unsuitable for quality testing. It is.

In the learning process, the number of image data to be learned by the AI is not particularly limited, but the larger the number, the higher the accuracy. In addition, in the learning process, each time image data of a test core that has been newly imaged and classified as a normal core or a defective core is added to the image data to be learned by the AI, the image data stored in the storage unit is added. The data may be used to retrain the AI.

In the learning process, a group of image data of normal cores and defective cores stored in the storage unit of the server is trained as learning data to generate a data group indicating a defective core, and the generated defective core is The data group shown is stored in the storage unit. In the learning process, the data group indicating a defective core generated by the AI is data that is compared with the test core to be determined in the analysis process.

The process management method for ground improvement work of the present invention includes an imaging step. In the imaging process, test cores to be evaluated are collected from the ground where the soil improvement work has been completed within 168 hours after the completion of the soil improvement work, and imaged at least 72 hours after the completion of the soil improvement work. This is the process of

In ground improvement work, a cement-based solidifying agent is sprinkled on the soft ground that is the target of ground improvement, and mixed with stirring, or by pumping a solidifying agent slurry into the ground that is the target of ground improvement. The ground improvement process is carried out by stirring and mixing with the soil inside. After the ground improvement process is completed, test cores are collected by boring in order to determine whether the ground improvement has been carried out normally. After collecting the test core to be determined, the collected test core is wrapped in a curing sheet or the like.

The time to collect test cores to be judged is selected as appropriate within 168 hours after the end of the ground improvement work. It is preferable to collect test cores from the ground to be improved between 72 and 96 hours after the completion of the ground improvement work, as this will prevent the test cores from being damaged when they are collected. . In addition, when collecting test cores, if the test cores can be collected from the ground to be improved without damaging them, the test should be carried out immediately after the completion of the ground improvement work or within 72 hours after the completion of the ground improvement work. cores can be collected. In addition, if the test core collected for quality testing after ground improvement work is found to be a defective core, the test core must be collected again from the ground targeted for ground improvement work. However, in that case, test cores for evaluation will be collected again within 168 hours after the completion of the ground improvement work.

In the imaging step, immediately before photographing the test core to be determined, the curing sheet surrounding the test core is peeled off to prevent drying, etc., and an image of the test core to be determined is captured.

In the imaging process, the timing for imaging the test core is at least 72 hours after the end of the ground improvement work, and is selected as appropriate between 72 and 168 hours after the end of the ground improvement work. . The imaging process includes, for example, (1) If the test core is collected between 72 and 96 hours after the end of the ground improvement work, immediately after the test core is collected, for example, within 48 hours after the test core is collected. An example of this is to take an image of the test core. Also, for example, (2) if the test core is collected within 72 hours after the end of the ground improvement work, the test core may be imaged after 72 hours or more have passed after the end of the ground improvement work. . In addition, for example, (3) a test core is collected between 72 and 96 hours after the completion of the ground improvement work, and the test core is collected immediately after collection, for example, within 48 hours after the test core is collected. If the core is determined to be defective as a result of taking an image, performing the imaging process, and then performing the transmission process, analysis process, and confirmation process, a test core is collected from the target ground again and the test core is imaged. There are things to do. In addition, for example, (4) a test core is collected within 72 hours after the completion of the ground improvement work, and the test core is imaged and the imaging process is performed after 72 hours or more have passed after the completion of the ground improvement work. If it is determined that the core is defective as a result of performing the transmission process, analysis process, and confirmation process, the test core may be sampled again from the target ground and imaged of the test core.

In the imaging step, the imaging device for imaging the test core is not particularly limited as long as it can obtain the captured image as digital data, and examples thereof include a digital camera and a smartphone with an imaging function. Then, by capturing an image of the test core with an imaging device, image data of the test core to be determined is obtained.

The process control method for the ground improvement process of the present invention includes a transmission process. The transmitting step is a step of transmitting image data of the test core to be determined obtained by performing the imaging step to the server.

In the transmission process, there are no particular restrictions on the method of transmitting the image data of the test core to be determined to the server, and it may be from a computer such as a desktop computer, mobile computer, or smartphone that is connected to the server via the Internet. , A method of directly transmitting the image data stored in the device to the storage section of the server, and a method of transmitting the image data to the storage section of the server by uploading the image data to the upload database using upload software. One method is to send the .

The quality control method for ground improvement work of the present invention includes an analysis step. In the analysis process, AI compares the image data of the test core to be determined received by the receiving section of the server with the data group indicating the defective core stored in the storage section, and determines whether the test core to be determined is the target test core. , is a step of determining whether or not the core corresponds to a defective core.

In the analysis step, the AI determines the degree of agreement between the test core to be determined and a data group indicating a defective core, and based on the degree of agreement, determines whether or not the test core corresponds to a defective core.

The process management method for ground improvement work of the present invention includes a confirmation process. The confirmation process is a process of confirming the analysis results in the analysis process. In the confirmation process, the analysis results from the analysis process are received from the transmitter of the server via the Internet on a computer such as a mobile computer or smartphone, and the results are uploaded to the upload database by the server. The person who obtained the analysis results can check the analysis results by receiving the analysis results on a computer such as a mobile computer or smartphone using the software for obtaining the uploaded analysis results. do.

Then, a confirmation step is performed, and if the analysis result shows that the test core is a normal core, the imaged test core is sent to a quality testing laboratory. On the other hand, if the analysis result is that the test core is a defective core, boring is performed on the ground where the ground improvement work has been performed again to obtain a new test core. For this reason, in the process control method for ground improvement work of the present invention, if the confirmation result is a defective core, re-sampling by boring can be performed at that point, so the ground improvement work can be completed. Compared to conventional quality control methods, where it is found that the test core needs to be re-collected in the preliminary test after 7 days or the main test after 28 days, this Less time wasted. In addition, in the process control method for ground improvement work of the present invention, it is possible to immediately confirm whether the test core is a normal core or a defective core after performing the imaging process, so whether the confirmation result is a defective core or not. In this case, re-sampling by boring can be carried out at that point, so test cores for the preliminary test 7 days later can be re-sampled within 7 days after the completion of the ground improvement work.

In the process control method for ground improvement work of the present invention, it is possible to determine whether or not a test core collected by boring is suitable for a quality test, without relying on the experience of a discriminator. In other words, the process control method for ground improvement work of the present invention has less variation than human judgment, and can reduce the number of erroneous judgments by judges, compared to the conventional method in which test cores were judged by humans. Compared to quality control methods, there is less time loss when cores for testing are improperly collected.

In the process control method for ground improvement work of the present invention, AI determines whether or not the test core collected by boring is suitable for quality testing, so there is no need to train a skilled judge.

The server used in the quality control method for ground improvement work of the present invention, and the AI and storage unit provided in the server are not particularly limited.

In the quality control method for ground improvement work of the present invention, the imaging step can be performed using an imaging jig or test core storage device described below.

[Imaging jig according to the first embodiment of the present invention]
The imaging jig according to the first aspect of the present invention is an imaging jig that is attached to a box containing test cores collected from the ground where soil improvement work has been performed, and includes:
a base having a storage section that is provided so as to span the inside and outside of one side wall of the box and can store the one side wall;
a clamp mechanism provided at the base and capable of clamping the one side wall in the thickness direction;
a shaft extending above the box from the base, configured to be extendable and retractable, and supporting an imager at the distal end;
a stabilizing rod extending along the upper surface of the one side wall in a direction penetrating the base and abutting the upper surface of the one side wall;
An imaging jig comprising:

A first embodiment of an imaging jig will be described with reference to FIGS. 1 to 3.
In the imaging jig 21 of the first embodiment, an imaging device 26 is attached to the box 11 for photographing the test core 12 taken from the ground that has undergone soil improvement work. A plurality of test cores 12 are stored in the box 11 .

As shown in FIG. 1, the box 11 includes a plurality of elongated slots 13 for storing a plurality of test cores 12, a peripheral wall 14 surrounding each slot 13, a bottom wall 15 defining the bottom of the slot 13, and a box 11. and a side wall 16 that surrounds the outermost peripheral portion of. The peripheral wall 14 has a pair of long walls 14A and a pair of short walls 14B. The box 11 has a rectangular shape when viewed from above. The side wall 16 of the box 11 has a long side 16A and a short side 16B. The slots 13 are provided side by side in the direction of the short side 16B. Each slot 13 houses an elongated cylindrical test core 12 . The longitudinal direction of the test core 12 is along the long side 16A direction of the box 11. The box 11 is usually made of wood, but may be made of resin or metal other than wood.

As shown in FIGS. 1 to 3, the imaging jig 21 includes a block-shaped base 22 provided so as to span the inside and outside of one side wall 16 of the box 11, and a storage section 23 provided in the base 22. , a clamp mechanism 24 provided on the base 22 and capable of clamping the one side wall 16 in the thickness direction; a shaft 25 extending upward from the base 22 to the box 11; and an imager 26 provided at the tip of the shaft 25. a first stabilizing rod 28 extending from the first surface 31 of the base 22 in a direction penetrating the base 22 along the upper surface of one side wall 16; A second stabilizing rod 29 extends along the upper surface of one side wall 16 in a direction penetrating the base 22.

The base 22 has a substantially trapezoidal block shape when viewed from the side. The base 22 may be constructed by connecting two metal plates with a joint such as a bolt. The base 22 has a first surface 31 and a second surface 32 opposite to the first surface 31. The base 22 has a storage section 23 into which one side wall 16 can be inserted and stored. The storage portion 23 is formed by cutting out a portion of the two plates that constitute the base portion 22 .

The shaft portion 25 is in the shape of a telescopic rod. The shaft portion 25 can support a support portion 27 and an imager 26 placed thereon at its tip. The shaft portion 25 can rotate around a fixed shaft 33 provided on the base portion 22 . As shown in FIG. 2, the shaft portion 25 can rotate between the first threaded portion 34 and the second threaded portion 35. The base portion 22 may have a through hole 36 formed as an arc-shaped elongated hole centered on the fixed shaft 33. The angle of the shaft portion 25 may be finely adjusted by passing a bolt or the like through the through hole 36 and supporting the shaft portion 25 with the bolt.

As shown in FIG. 1, the support shaft portion 25 has a movable claw portion that can hold the image pickup device 26 therebetween. The support portion 27 is fixed to the shaft portion 25 via a universal joint or the like, and the inclination of the image pickup device 26 with respect to the box 11 can be finely adjusted. Imager 26 may be a smartphone. The image pickup device 26 is not limited to a smartphone, and may be another type of image pickup device such as a compact digital camera or a film-type compact camera.

As shown in FIG. 2, the storage portion 23 is formed as a notch that is provided so as to penetrate the first surface 31 and the second surface 32. As shown in FIG. One side wall 16 can be stored inside the storage part 23.

The clamp mechanism 24 includes a pair of insertion rods 38 that are inserted into the inside of the box 11, and a cylinder portion 41 that is provided outside the box 11 and presses one side wall 16. The pair of insertion rods 38 are provided in the base 22 at positions spaced apart from each other in the thickness direction. Each of the insertion rods 38 extends in the vertical direction and is arranged along the inner surface of one side wall 16 of the box 11. The tip of the insertion rod 38 may come into contact with the bottom wall 15 of the box 11. It is preferable that the insertion rod 38 and the cylinder part 41 be formed of a metal material or the like.

The cylinder portion 41 is configured as a screw-type cylinder, and can be moved toward or away from one side wall 16 by rotating the screw portion. The clamp mechanism 24 can clamp one side wall 16 by sandwiching the one side wall 16 between the insertion rod 38 and the cylinder part 41. The base 22 is stably fixed to one side wall 16 of the box 11 by a clamping mechanism 24 . The cylinder portion 41 is preferably formed of a metal material or the like.

The first stabilizing rod 28 is provided adjacent to the edge of the storage portion 23 . Although the first stabilizing rod 28 is fixed to the base 22 by pinching the edge of the storage section 23, the method of fixing the first stabilizing rod 28 to the base 22 is not limited to this. do not have. The first stabilizing rod 28 may be fixed to the base 22 by welding, insertion into a through hole, or the like. The first stabilizing bar 28 extends laterally and can abut the upper surface of one side wall 16 . The first stabilizing rod 28 is preferably formed of a metal material or the like.

The second stabilizing rod 29 is provided so as to protrude in the opposite direction to the first stabilizing rod 28. The second stabilizing rod 29 is provided adjacent to the edge of the storage portion 23 . The method of fixing the second stabilizing bar 29 is the same as the method of fixing the first stabilizing bar 28. The second stabilizing bar 29 extends laterally and can abut the upper surface of one of the side walls 16 . The second stabilizing rod 29 is preferably formed of a metal material or the like. Note that in this embodiment, the first stabilizing rod 28 and the second stabilizing rod 29 are configured separately, but the configuration of the stabilizing rod is not limited to this. By providing one stabilizing rod so as to penetrate the base 22 and bringing this one stabilizing rod into contact with the upper surface of one side wall 16, the roles of the first stabilizing rod 28 and the second stabilizing rod 29 are reduced. may be carried by one stabilizing rod.

The operation of the imaging jig 21 of this embodiment will be explained. The imaging jig 21 of this embodiment can be easily fixed to the box 11 by the clamp mechanism 24. Furthermore, since the clamp mechanism 24 firmly fixes the image pickup device 26 to the box 11, the image pickup device 26 will not be shaken in the thickness direction of the side wall 16. Furthermore, since the inclination of the base 22 is stably maintained by the first stabilizing rod 28 and the second stabilizing rod 29, the image pickup device 26 does not shake in the direction in which the side wall 16 extends. As a result, the test core 12 can be photographed without blurring, and a high-resolution image of the test core 12 can be obtained.

According to this embodiment, the following can be said. The imaging jig 21 is an imaging jig that is attached to the box 11 that houses the test core 12 collected from the ground where the ground improvement work has been performed. a base 22 having a storage section 23 which is provided so as to extend over the base 22 and which can accommodate one side wall 16; a clamp mechanism 24 which is provided in the base 22 and which can clamp one side wall 16 in the thickness direction; A shaft part 25 extends above the base part 22 and extends above the base part 22 and extends in a direction penetrating the base part 22 along the upper surface of one side wall 16 . a stabilizing rod that abuts the top surface.

According to this configuration, since the clamp mechanism 24 can securely fix the base 22 to the side wall 16 in the thickness direction of the side wall 16, the image pickup device 26 does not shake in the thickness direction of the side wall 16 during imaging. In addition, since the stabilizing rod can come into contact with the upper surface of the side wall 16 in the direction in which the side wall 16 extends (the direction passing through the base 22), the imager 26 does not shake in the direction in which the side wall 16 extends during imaging. There is no need to put it away. These structures prevent the image pickup device 26 from shaking during image capture, and it is possible to capture high-resolution images.

In this case, stabilizing bars are provided along the top surface of the one side wall 16 on both sides with the base 22 in between. According to this configuration, since stabilizing rods are arranged on both sides of the base 22, stability during imaging can be further improved. This makes it possible to capture images with higher resolution.

[Imaging jig according to the second embodiment of the present invention]
The imaging jig according to the second embodiment of the present invention is attached to a box in which a plurality of slots are arranged side by side in the short side direction to house a plurality of test cores taken from the ground where soil improvement work has been performed. An imaging jig,
a turret facing the box and to which an imager is fixed;
a plurality of joints provided on the turret;
A plurality of legs extending from the plurality of joints are configured to be rotatable in the short side direction of the box with respect to the turret via the plurality of joints, and are configured to be rotatable in the short side direction of the box through the plurality of joints. By being inserted into the inside of the peripheral wall, it can stand up against the box, and by being inserted into the inside of the peripheral wall of the slot located on the outside in the short side direction, the distance between the turret part and the box is shortened, a plurality of legs that can be inserted into the inner side of the peripheral wall of the slot located on the inner side in the short side direction to increase the distance between the turret part and the box;
An imaging jig comprising:

A second embodiment of the imaging jig 51 will be described with reference to FIGS. 4 to 12.
In the imaging jig 51 of the second embodiment, an imaging device 26 is attached to the box 11 for photographing the test core 12 taken from the ground that has undergone soil improvement work. A test core 12 is housed in the box 11 . The shapes of the box 11 and the plurality of test cores 12 are the same as in the first embodiment.

As shown in FIG. 4, the imaging jig 51 includes a tower section 52 facing the box 11 and to which the imager 26 is fixed, a plurality of joint sections 53 provided on the tower section 52, and a plurality of joint sections. A plurality of legs 54 extend from 53 toward the box 11.

The turret part 52 has a frame part 55 formed by framing a rod-shaped frame into a square so as to form a rectangle when viewed from above, and a flat plate-shaped mounting part 56 fixed on the frame part 55. . It is preferable that the frame portion 55 is formed of a metal material or the like.

As shown in FIGS. 6 to 8, the mounting section 56 includes a mounting plate 57 provided with a through hole 57A, and a pair of slide pieces 58 that can slide on the mounting plate 57. The image pickup device 26 can be placed on the placement plate 57 . The imager 26 has a configuration similar to that of the first embodiment. Each of the pair of slide pieces 58 has a substantially "L"-shaped cross section. Each of the pair of slide pieces 58 can slide (move forward and backward) on a rail (not shown) formed on the mounting plate 57. The pair of slide pieces 58 can hold the imager 26 (smartphone) between them.

As shown in FIG. 7, the installation position of the imager 26 is arbitrary, and the imager 26 can be placed at any position on the mounting plate 57. The pair of slide pieces 58 can be moved appropriately according to the position of the image pickup device 26. The imager 26 is mounted on the mounting plate 57 so that the lens portion 26A matches the through hole 57A. It is preferable that the mounting plate 57 and the pair of slide pieces 58 are formed of a metal material or the like.

The mounting section 56 may be configured to be slidable relative to the frame section 55. The sliding mechanism between the placing part 56 and the frame part 55 can take various embodiments. For example, the mounting section 56 may be attached to the frame section 55 in the manner shown in FIG. In this example, the mounting section 56 is movable in the X direction and the Z direction shown in FIG. 5 with respect to the frame section 55, but is fixed in the Y direction.

Moreover, the mounting part 56 may be attached to the frame part 55 in the manner shown in FIG. In this example, the mounting section 56 is movable in the X direction shown in FIG. 5 with respect to the frame section 55, but is fixed in the Y direction and the Z direction.

Moreover, the mounting part 56 may be attached to the frame part 55 in the manner shown in FIG. In this example, the mounting section 56 is movable in the X direction and the Y direction shown in FIG. 5 with respect to the frame section 55, but is fixed in the Z direction.

Moreover, the mounting section 56 may be attached to the frame section 55 in the manner shown in FIG. 11. In this example, the mounting portion 56 is fixed to the frame portion 55 by an adhesive such as welding. In this example, the placing section 56 cannot be moved with respect to the frame section 55.

It is preferable that the plurality of legs 54 are constituted by rod-shaped pipes. The plurality of legs 54 can rotate along the short side direction of the box 11 as shown by arrows in FIG. 4 by the action of the plurality of joints 53. The plurality of legs 54 are inserted into the inner side of the peripheral wall of the slot 13 of the box 11 and abut against the peripheral wall 14 and the bottom wall 15. This prevents the plurality of legs 54 from expanding outward in the short side direction and maintains the upright state. It is preferable that the plurality of legs 54 be formed of a metal material or the like.

The form of the plurality of legs 54 is not limited to one extending in the vertical direction as shown in FIG. 4. For example, as shown in FIG. 12, it may extend obliquely to the bottom wall 15 of the box 11.

The operation of the imaging jig 51 of this embodiment will be explained. In this embodiment, the installation position and installation height of the image pickup device 26 can be changed by the following method. The user can, for example, slide the mounting section 56 with respect to the frame section 55 so as to directly face the portion of the test core 12 that is desired to be photographed. Therefore, in this embodiment, the image pickup device 26 can be freely moved in the plane direction formed by the frame portion 55.

Furthermore, in this embodiment, the installation height of the image pickup device 26 can be changed by changing the positions of the plurality of legs 54. For example, by rotating the plurality of legs 54 outward in the short side 16B direction by the action of the joint part 53 and inserting them into the inside of the peripheral wall 14 of the slot 13 located on the outside in the short side 16B direction, the turret part 52 The distance between the camera and the box 11 can be shortened (the installation height of the imager 26 can be lowered). Alternatively, by rotating the plurality of legs 54 inward in the direction of the short side 16B by the action of the joint part 53 and inserting them into the inside of the peripheral wall 14 of the slot 13 located on the inside in the direction of the short side 16B, the turret part 52 and the box 11 can be increased (the height of the image pickup device 26 can be increased). In this case, the range indicated by the arrow W in FIGS. 4 and 12 is the movable range of the leg portion 54 (the range in which the leg portion 54 can be installed).

In this manner, in this embodiment, the position of the image pickup device 26 can be arbitrarily changed with respect to the plane direction and the height direction formed by the frame portion 55. Therefore, according to the imaging jig 51 of the present embodiment, by appropriately changing the installation position of the imager 26, it is possible to, for example, take a close-up of a desired part of the test core 12, or It is possible to photograph the entire image of the core 12, and it is possible to realize an imaging jig 51 that is easy for the user to use.

According to this embodiment, the following can be said. The imaging jig 51 is an imaging jig that is attached to a box 11 in which a plurality of slots 13 are arranged side by side in the short side direction to house a plurality of test cores 12 taken from the ground where soil improvement work has been performed. 51, a tower part 52 facing the box 11 and to which the imager 26 is fixed, a plurality of joint parts 53 provided on the tower part 52, and a plurality of leg parts 54 extending from the plurality of joint parts 53. The box 11 is configured to be rotatable in the short side direction of the box 11 with respect to the turret part 52 via the plurality of joints 53, and is inserted into the inner side of the peripheral wall 14 of the plurality of slots 13. By inserting it into the inside of the peripheral wall 14 of the slot 13 located on the outside in the direction of the short side 16B, the distance between the turret part 52 and the box 11 is shortened, and the slot located on the inside in the direction of the short side 16B It is provided with a plurality of leg parts 54 which can lengthen the distance between the turret part 52 and the box 11 by being inserted into the inside of the peripheral wall 14 of 13.

According to this configuration, the installation height of the image pickup device 26 can be changed depending on the installation position of the leg portion 54, and it is possible to provide the imaging jig 51 that is easy to use for the user.

[Test core storage device of the present invention]
The test core storage device of the present invention is a rectangular box that stores cylindrical test cores collected from the ground where soil improvement work has been performed and wrapped in a curing bag. a box having a short wall, and each of the pair of short walls is provided with a recess;
a dish portion having an arcuate cross section and provided between the pair of short walls and on which the test core is placed;
a pair of rotating shafts extending from the dish toward the recess and rotatably supported within the recess;
A slider that runs on the pair of long walls so as to move along the longitudinal direction of the test core placed on the dish, and a curing bag that is fixed to the slider and wraps the test core are cut. a cutter portion having a cutting blade;
This is a testing core storage device equipped with:

An embodiment of the test core storage device 61 will be described with reference to FIGS. 13 to 19. The test core storage device 61 is capable of storing test cores 12 collected from ground that has undergone soil improvement work in the box 11 without damaging them. The form of the box 11 is substantially the same as the first form, but differs only in having a recess 62 in the short wall 14B. Further, in this embodiment, the test core 12 is collected from the ground and then wrapped in a curing bag 63, but other than that, the test core 12 is the same as the first embodiment. As shown in FIG. 13, the recess 62 is formed in the shape of a groove that falls downward from the upper surface of the short wall 14B.

As shown in FIGS. 14, 15, and 17, the test core storage device 61 is provided between the box 11 having the above structure and a pair of short walls 14B, and has a dish portion on which the test cores 12 are placed. 64, a pair of rotating shafts 65 extending from the dish part 64 toward the recess 62, a disc-shaped handle part 66 provided at the end of the rotating shaft 65 on the opposite side to the end on the dish part 64 side, and the dish part 64. A cutter section 67 that moves along the longitudinal direction of the test core 12 placed on the section 64 and a gap section 68 provided at a position between the dish section 64 and the bottom wall 15 of the box 11 are provided.

As shown in FIG. 15, the dish portion 64 is formed to have an arcuate cross section. It is preferable that the dish portion 64, the rotating shaft 65, and the handle portion 66 be integrally formed of a metal material. These are preferably joined together by welding or the like. The rotating shaft 65 is inserted into the recess 62 and rotatably supported within the recess 62 by the recess 62 .

As shown in FIG. 17, the cutter section 67 includes a slider 71 that runs on the pair of long walls 14A, and a cutting blade 72 that is fixed to the slider 71 and cuts the curing bag 63 that wraps the test core 12. As shown in FIG. 12, the slider 71 includes, for example, a grip portion 71A that is gripped by a user, and a roller portion 71B that is provided at the end of the grip portion 71A and that can travel on the long walls 14A while coming into contact with a pair of long walls 14A. and has. The form of the slider 71 is arbitrary, and as long as it can slide in the longitudinal direction with respect to the test core 12, it may be a slider of another type. A cutting blade 72 is supported and fixed at the center of the grip portion 71A.

The operation of the testing core storage device 61 of this embodiment will be explained. As shown in FIG. 16, the test core 12 taken from the ground that has undergone soil improvement work is placed on the upper side of the dish portion 64. At this time, the test core 12 is wrapped in a curing bag 63 (a vinyl bag for curing). The dish portion 64 on which the test core 12 is placed in this manner is installed in the slot 13 of the box 11, as shown in FIG. At this time, the rotating shaft 65 is disposed within the recess 62, and the handle portion 66 is disposed outside the pair of short walls 14B of the box 11.

Subsequently, as shown in FIG. 17, the cutter section 67 is installed in the box 11. The cutter section 67 is installed so that its roller section 71B comes into contact with the upper surface of the pair of long walls 14A of the box 11. The cutter section 67 installed in this manner is slidable along the longitudinal direction of the test core 12 placed on the dish section 64 .

When the user grips the gripping portion 71A of the cutter portion 67 and slides the cutter portion 67 along the longitudinal direction of the test core 12 by a length corresponding to the entire length of the test core 12, the cutting blade 72 The curing bag 63 is cut by traveling on the protective core 12.
Furthermore, the user installs a curing sheet 73 different from the curing bag 63 in the gap 68, as shown in FIG. In this state, the user rotates the handle part 66 to rotate the dish part 64 by about 90 to 180 degrees, and drops the test core 12 onto the bottom wall 15 of the box 11. At that time, the cut curing bag 63 is removed from the test core 12.

In this way, the test core 12 is stored in the box 11 and the curing bag 63 is peeled off, and the work of storing the test core 12 in the box 11 is completed. After the test core 12 is stored in the box 11, an image of the test core 12 is taken using an imaging device.

After imaging the test core 12, as shown in FIG. 19, the test core 12 placed on the bottom wall 15 of the box 11 is wrapped in a curing sheet 73 and sent to the testing site.

According to this embodiment, the following can be said. The test core storage device 61 is a rectangular box 11 that stores cylindrical test cores 12 collected from the ground where soil improvement work has been performed and wrapped in a curing bag 63, and includes a pair of long walls 14A, A box 11 having a pair of short walls 14B and having a recess 62 in each of the pair of short walls 14B, and a test core 12 having an arc-shaped cross section and provided between the pair of short walls 14B. A pair of rotation shafts 65 extending from the tray portion 64 toward the recess 62 and rotatably supported within the recess 62; A slider 71 that runs on a pair of long walls 14A so as to move along the longitudinal direction of the test core 12 placed thereon, and a cutting blade that cuts the curing bag 63 that is fixed to the slider 71 and that wraps the test core 12. 72, and a cutter section 67 having the following.

According to this configuration, the test core 12 can be stored in the box 11 without damaging it, and the curing bag 63 wrapped around the test core 12 can be easily removed by the cutter section 67.

In this case, the test core storage device 61 includes a disk-shaped handle portion 66 provided at the end of the rotating shaft 65 opposite to the end on the dish portion 64 side. According to this configuration, since the handle part 66 is connected to the dish part 64, the user can easily perform an operation of rotating the dish part 64. Thereby, a more user-friendly test core storage device 61 can be realized.

In this case, the test core storage device 61 is a gap section 68 provided between the tray section 64 and the bottom wall 15 of the box 11, into which a curing sheet 73 for packaging the test cores 12 is inserted. A possible gap 68 is provided. According to this configuration, after the test core 12 is housed in the box 11 and imaged, a curing sheet 73 different from the curing bag 63 can be wrapped around the test core 12. As a result, after imaging, the test core 12 can be quickly wrapped in the curing sheet 73 without having to take the test core 12 out of the box 11, and work efficiency can be improved.

Then, using the imaging jig of the present invention, an imaging process related to the process control method for ground improvement work using artificial intelligence (AI) of the present invention is performed. That is, the imaging jig of the present invention is installed in a box that stores the test core, an imaging device is fixed to the imaging jig of the present invention, and the imaging device images the test core to be determined. Accordingly, the imaging step of the process management method for ground improvement work using artificial intelligence (AI) of the present invention is performed.

Moreover, the imaging process related to the process control method for ground improvement work using artificial intelligence (AI) of the present invention is performed using the testing core storage device of the present invention. That is, the test core to be determined wrapped in a curing bag is placed on the tray of the test core storage device of the present invention, the curing bag is cut by the cutter, and the pan is rotated. At the same time, the curing bag is peeled off from the test core to be determined, and the test core to be determined is stored in a box for storing test cores.Then, the imaging device is used to image the test core to be determined. By doing so, the imaging step of the process management method for ground improvement work using artificial intelligence (AI) of the present invention is performed.

Furthermore, before placing the test core to be determined wrapped in a curing bag on the tray portion of the test core storage device of the present invention, the test core to be determined is placed in the storage portion of the box that stores the test core to be determined. , a curing sheet is laid down, and then an imaging process is performed according to the process control method for ground improvement work using artificial intelligence (AI) of the present invention, and then the test core to be judged from which the curing bag has been removed is placed in the curing sheet. To wrap.

The process control system for ground improvement work using artificial intelligence (AI) of the present invention is a ground improvement work using artificial intelligence (AI) for carrying out the process control method for ground improvement work using artificial intelligence (AI) of the present invention. process control system,
At least an image of the test core to be determined, taken from the ground where the ground improvement work was performed, within 168 hours after the end of the ground improvement work, and taken at least 72 hours after the end of the ground improvement work. It has a transmitting device that transmits data to a server, and a server that connects to the transmitting device via the Internet,
The server is
a receiving unit that receives image data of the test core to be determined sent from the transmitting device;
A data group of images of normal cores suitable for quality testing and defective cores not suitable for quality testing, collected within 168 hours after the completion of soil improvement work and taken at least 72 hours after the completion of ground improvement work. A memory unit that stores information,
Normal cores and quality test that are stored in the storage unit of the server and are collected within 168 hours after the end of the ground improvement work and are imaged at least 72 hours after the end of the ground improvement work and are suitable for quality testing. learns a group of image data of defective cores that are not suitable as learning data, generates a group of data indicating defective cores, stores the generated data group indicating defective cores in the storage unit, and a receiving unit of the server; Compares the image data of the test core to be determined received with the data group indicating a defective core stored in the storage unit, and determines whether or not the test core to be determined corresponds to a defective core. Artificial intelligence (AI) that determines whether
to have
This is a process management system for ground improvement work that uses artificial intelligence (AI).

The process management system for ground improvement work using artificial intelligence (AI) of the present invention is a system for performing the process control method for ground improvement work of the present invention.

The process management system for ground improvement work using artificial intelligence (AI) of the present invention transmits at least the image data of the test core to be determined, which is collected from the ground where the ground improvement work has been carried out and photographed, to the server. The computer has a transmitting device, for example, an electronic computer such as a desktop personal computer, a mobile personal computer, a smartphone, etc., and a server connected to the transmitting device via the Internet.

The server is
a receiving unit that receives image data of the test core to be determined sent from the transmitting device;
A data group of images of normal cores suitable for quality testing and defective cores not suitable for quality testing, collected within 168 hours after the completion of soil improvement work and taken at least 72 hours after the completion of ground improvement work. A memory unit that stores information,
Normal cores suitable for quality testing and imaged at least 72 hours after the end of ground improvement work, stored in the storage unit of the server, taken within 168 hours after the end of the ground improvement work and The image data group of unsuitable defective cores is learned as learning data, a data group indicating defective cores is generated, the generated data group indicating defective cores is stored in a storage unit, and the data group is received by a receiving unit of the server. Artificial intelligence that compares the image data of the test core to be determined with a data group indicating a defective core stored in the storage unit and determines whether the test core to be determined corresponds to a defective core. (AI) and
Equipped with

11 Box 12 Test core 13 Slot 14 Peripheral wall 14A Long wall 14B Short wall 15 Bottom wall 16 Side wall 16A Long side 16B Short side 21 Imaging jig 22 Base 23 Storage part 24 Clamp mechanism 25 Shaft 26 Imager 26A Lens part 27 Support Section 28 First stabilizing rod 29 Second stabilizing rod 31 First surface 32 Second surface 33 Fixed shaft 34 First threaded section 35 Second threaded section 36 Through hole 38 Insertion rod 41 Cylinder section 51 Imaging jig 52 Turret part 53 Joint part 54 Leg part 55 Frame part 56 Placement part 57A Through hole 57 Placement plate 58 Slide piece 61 Test core storage device 62 Recessed part 63 Curing bag 64 Dish part 65 Rotating shaft 66 Handle part 67 Cutter part 68 Gap Part 71 Slider 71A Gripping part 71B Roller part 72 Cutting blade 73 Curing sheet

Claims (2)

Artificial intelligence (AI) uses quality test images that are stored in the server's storage and taken within 168 hours after the end of the ground improvement work, and taken at least 72 hours after the end of the ground improvement work. A data group indicating a defective core is generated by learning a group of image data of a normal core that is suitable and a defective core that is not suitable for a quality test as learning data, and the generated data group indicating a defective core is stored in the storage unit. learning process to
Imaging in which an image of the test core to be judged is taken from the ground where the ground improvement work has been performed, within 168 hours after the end of the ground improvement work, and is taken at least 72 hours after the end of the ground improvement work. process and
a transmission step of transmitting image data of the test core to be determined obtained through the imaging step to the server;
Artificial intelligence (AI) compares the image data of the test core to be determined, received by the reception unit of the server, with the data group indicating the defective core stored in the storage unit, and determines the target to be determined. an analysis step of determining whether or not the test core corresponds to a defective core;
a confirmation step of confirming the analysis results in the analysis step;
to have,
A process management method for ground improvement work using artificial intelligence (AI). A process control system for ground improvement work using artificial intelligence (AI) for carrying out the process control method for ground improvement work using artificial intelligence (AI) according to claim 1,
At least an image of the test core to be determined, taken from the ground where the ground improvement work was performed, within 168 hours after the end of the ground improvement work, and taken at least 72 hours after the end of the ground improvement work. It has a transmitting device that transmits data to a server, and a server that connects to the transmitting device via the Internet,
The server is
a receiving unit that receives image data of the test core to be determined sent from the transmitting device;
A data group of images of normal cores suitable for quality testing and defective cores not suitable for quality testing, collected within 168 hours after the completion of soil improvement work and taken at least 72 hours after the completion of soil improvement work. A memory unit that stores information,
Normal cores and quality test that are stored in the storage unit of the server and are collected within 168 hours after the end of the ground improvement work and are imaged at least 72 hours after the end of the ground improvement work and are suitable for quality testing. learns a group of image data of defective cores that are not suitable as learning data, generates a data group indicating defective cores, stores the generated data group indicating defective cores in the storage unit, and a receiving unit of the server; Compares the image data of the test core to be determined that is received with the data group indicating a defective core stored in the storage unit, and determines whether or not the test core to be determined corresponds to a defective core. Artificial intelligence (AI) that determines whether
to have
A process management system for ground improvement work that uses artificial intelligence (AI).

Images