Digital Extraordinary Administrative Advisory Committee Working Group (2nd)

Overview

Date and Time: Tuesday, February 22, 2022 (2022) from 10:00 to 12:00
Location: Online
Agenda:
1. Opening
2. Proceedings
  1. Hearings with companies that provide digital technologies that can be used for regulation reviews and relevant ministries and agencies that have conducted preliminary regulation reviews
  2. Exchange of opinions
3. Adjournment

Materials

Related Information

The 2nd Digital Ad Hoc Working Group was held. We learned advanced technologies to replace analog regulation through public-private cooperation.

Minutes, etc.

Date

Tuesday, February 22, 2022 (2022), from 10:00 to 12:00

Location

Online Meetings

Attendees

Chairman

Su Beng Seiichi Kobayashi

Members

Junji Annen (Attorney-at-law, Professor of the Graduate School of Law
Tatsuhiko Inadani (Professor, Graduate School of Law, Kyoto University)
Katsuya Uenoyama (President of PKSHA Technology, Inc.)
Takafumi Ochiai (Attorney at law, Atsumi & Sakai, Foreign Law Joint Enterprise)
Katsunori Nemoto (Senior Managing Director, Japan Business Federation)
Masakazu Masushima (Attorney-at-Law, Mori Hamada & Matsumoto)

Minutes

Mr. Suga: We would like to hold the third meeting of the Working Group on Now that it is time, I would like to open the second session of the "Digital Extraordinary Administrative Advisory Committee Working Group." Thank you very much.
Today, all members are participating online. Member Sugawara will be absent due to personal reasons, and Member Uenoyama will leave the meeting in the middle of the meeting.
First of all, I would like to ask for an address from Senior Vice-Minister for Digital Affairs Kobayashi, who is the chair of this working group, prior to the opening of the second working group.

Senior Vice-Minister for Digital Affairs Kobayashi: Good morning, everyone from . Members of the Working Group, and today, people from the Ministry of Land, Infrastructure, Transport and Tourism, as well as people from various private sector companies, are participating.

The working group of the Digital Rincho is a forum to check each ministry and agency and Digital Extraordinary Administrative Advisory Committee based on the digital principles compiled in law, discuss how much it can be reflected in accordance with the digital principles, and correct them together.

I believe that there are two points that need to be addressed in order to move forward in a more concrete manner.

One is the extent to which so-called regulations on paper and in-person processes, which is currently limited to analog means, can be automated. In other words, I think it is almost synonymous with the extent to which existing technology has advanced and can be used for social digitalization. We need to know this ourselves, and I think that everyone in each ministry and agency will know it. If private sector's technology has advanced so much, they will feel that the rules can be changed. I think that everyone should understand it. implementation

The second point is that if such technology exists in private sector, in order to implementation it, it is necessary to verify the technology, such as safety inspections, depending on the field. If we say that it takes three years to verify the technology, it will take a long time to reform. So, how quickly can we verify the technology and embed it in this reform? And since inspections and face-to-face meetings will be conducted online in society, I believe that each ministry and agency must determine whether the implementation is possible and make it practical for each local government.

To this end, we have an image that it would be good if it were like a catalog if possible. In fact, the Ministry of Land, Infrastructure, Transport and Tourism has already started cataloging. Today, the Ministry of Land, Infrastructure, Transport and Tourism is leading the way in doing what we want to do, and we would like to know what kind of know-how we have and how we can do it quickly. In addition, we would like to hear from related companies about current technology trends. Today, we would like to hear about telecommunications technology recommended by the Ministry of Economy, Trade and Industry, and we would like to have the two points I mentioned earlier firmly imaged within us and utilize them in discussions with each ministry and agency. Thank you very much for your cooperation today.

Vice-Chairman: Thank you very much, Mr. .

I am Mr. Yasune, Vice Chairman. I will be your moderator.

I would like to start the proceedings immediately. Today's proceedings are scheduled to be heard from the Road Bureau of the Ministry of Land, Infrastructure, Transport and Tourism and four companies that possess advanced technologies: KOKUSAI KOGYO CO.,LTD., Photon Lab Co., Ltd., CHODAI CO.,LTD., and Grid Skyway LLC.

First of all, I would like to have a presentation from Mr. Masanori Wakao, Director of the Technology Planning Office, National Highway and Technology Division, Road Bureau, Ministry of Land, Infrastructure, Transport and Tourism. Thank you very much.

Director Wakao: Regarding the

On my part, I would like to explain about the promotion of the introduction of new technologies for the inspection of road structures as "promotion of the introduction of new technologies for road infrastructure maintenance."

I will explain it according to the material. Please take a look at page 1.

On page 1, we are conducting statutory inspections of road structures. In December 2012, as you may recall, the ceiling plates of the Sasago Tunnel fell on the Chuo Expressway, which caused many deaths.

Based on this, we, the Road Bureau, must thoroughly inspect road structures, and in June 2013, the law was revised to stipulation inspection standard.

I would like to skip this step. Since 2014, the first round of regular inspections has been conducted once every five years. Since 2019, the second round has been conducted, and this is the third year of the second round.

On the next page is the legal system for the standards. In the "Standard System for law Inspections" below, as I mentioned earlier, the Cabinet Order stipulates that maintenance, inspections, and measures will be taken, and the Ministerial Ordinance states that the inspection will be conducted once every five years on the basis of a nearby visual inspection. In addition, the Notice states that the results of the soundness diagnosis will be divided into four stages, but as with other legal systems, minimum regulations have been established here, and the Road Bureau provides technical advice in the form of periodic inspection guidelines on how to actually conduct inspections.

It is the third page.

As I mentioned earlier, the first round of inspections was conducted from 2014 to 2018, and the first round was based on the neighboring visual inspection. However, amid the progress of various technologies, there is debate over whether technologies that can complement, substitute, and enhance the neighboring visual inspection should also be utilized.

Under such circumstances, the committee of experts has decided on how to respond, and we will revise the Periodic Inspection Guidelines that I mentioned earlier. Before the second round, in the [Points to Note for law Operation] written in the middle square on the right side, we have added a part written in red, saying, "A person who conducts a periodic inspection must grasp the status of the visual inspection by using a nearby visual inspection, or by using a method that is judged to provide information that can perform a diagnosis of soundness equivalent to that obtained by using his or her own nearby visual inspection." We have made it so that the status of the runway can be grasped by a method other than the nearby runway.

However, some people say that it is difficult for an inspector to make an individual judgment on the point that "it is possible to make a diagnosis of soundness equivalent to that made by one's own neighboring visual inspection." Therefore, we have created the "Guidelines for the Use of New Technology" and the "Inspection Support Technology Performance Catalog" to make a systematic judgment on this point.

Then, on page 4, what kind of new technologies are there? The upper side is the inspection we used to do, and the lower side is the inspection using inspection support technology.

On the far left side, we use image measurement, where places that are extremely difficult to see in visual inspection are photographed by drones, monitoring using sensors, image measurement in tunnels, and non-destructive inspection, where damage is confirmed by laser and other technologies.

Under the previous section, we have created guidelines that can be used as a reference in order to utilize new technologies for inspections. These guidelines are for road administrators such as local government and the Ministry of Land, Infrastructure, Transport and Tourism. The contractors are consultants and the like, and we have them do the work of inspection work under a consignment contract. When we entrust the work, this example shows the process of how to introduce new technologies for inspection.

In addition, as I will explain later, we have compiled a catalog, and this guideline shows how to check how to use the catalog between the supplier and the supplier.

The next page 6 is the inspection support technology performance catalog.

This performance catalog was formulated in February 2019. At the time of formulation, there were only 16 technologies, but it is revised every year. By 2020, there were 80 technologies, and at present, 131 technologies are listed in this catalog. By referring to this catalog, we will promote the use of new technologies for inspection.

This catalog is divided into four items at the bottom. As you can see on the right, image measurement is the technology to measure by drones and laser scanning, which is 34 technologies for bridges and 16 technologies for tunnels. Non-destructive inspection is the technology to grasp internal damage by electromagnetic waves and radar, which is 19 technologies for bridges and 13 technologies for tunnels. Measurement and monitoring is the technology to monitor displacement and such by sensors, which is 38 technologies for bridges and 8 technologies for tunnels. Data collection and communication is the same for bridges and tunnels, but it is a technology item with three technologies.

Regarding the procedures for expanding these technologies in the performance catalog, first of all, we will consider what technologies will be subject to the public offering, and then we will invite proposals in the form of a technology public offering. Then, we will conduct a field testing for the technologies that have been invited, and based on the validation of the field testing, we will expand the performance catalog and publish it in the performance catalog.

The above is the flow of the bridge, and the tunnel is in exactly the same situation, so I will not explain about this.

It is very simple, but this is the explanation from the Road Bureau about the promotion of new technology introduction.

Vice-Chairman: Thank you very much, Mr. , Director of the Office, thank you very much.

First of all, I would like to ask one question that will be the premise of the discussion. According to your explanation, I saw the text "Realizing efficiency and sophistication of periodic inspections" published in the Inspection Support Technology Performance Catalog, which is exactly the catalog you mentioned. Is it possible to make visual inspection inspections of infrastructure facilities such as tunnels completely unmanned by using the technology published in the catalog?

Director Wakao: Regarding the .

I believe it will be possible if we can have all the functions that can be substituted in terms of technology. However, at present, for example, there are various types and types of bridges, and for tunnels, for example, there are places where it is impossible to confirm details with current technology unless it is done in visual inspection, such as the back side of a jet fan. If future technological innovation makes it possible to determine such places with technology equivalent to that of visual inspection, I believe it will be possible to do so.

We are checking such new technologies in Field testing and other places where we are currently working, to confirm how accurate they are, and then putting them on the catalog.

Vice-Chairman: Thank you very much, Mr. .

If there are any opinions or questions from the members regarding your presentation, please let me know.

Mr. Ochiai, please.

Ochiai Member: Thank you, . My name is Ochiai.

I would like to express my gratitude to the Ministry of Land, Infrastructure, Transport and Tourism for taking up this theme at regulatory reform's Growth Strategies Working Group and taking the lead in promoting it.

I would like to ask two questions. One is, in many cases, there are regulations such as notifications to enable the use of such technologies. What kind of descriptions are used to enable the use of such technologies?

The second point is about how to write and publish this catalog. Some people, such as local government, say that it is difficult to understand what technology can be used, so I think they are compiling this. In that case, if there are any points that you think we should be careful about when other ministries and agencies develop them, please tell us. Please tell us these two points.

Vice-Chairman: Thank you very much, Mr. , Director of the Office, what do you think?

Director Wakao: Regarding the First, regarding the first point, it is in parentheses in the middle of the right column on page 3. As I explained earlier, originally, regarding the periodic inspection, the ministerial ordinance above the first round on the left side says, "In principle, the periodic inspection shall be conducted once every five years depending on the neighboring visual inspection." However, we have received opinions from road administrators that it is quite difficult to use the new technology even though it says that the new technology will be used in areas other than the neighboring visual inspection. In the second round, we left this ministerial ordinance as it is, and in the points of attention of the periodic inspection guidelines, which are used as a reference by road administrators to conduct periodic inspections, we said, "A method in which it is determined that information can be obtained by the neighboring visual inspection or can be used to diagnose the soundness equivalent to that obtained by one's own neighboring visual inspection." This is a method that uses new technology, and by revising these guidelines, each road administrator can use the new technology.

The second point is something that we should be careful about when creating a catalog. One is that if we do not secure performances based on certain standards, people from various private sector will apply for this catalog. So, we will have certain standards so that we do not choose a catalog arbitrarily. For example, we need to check the cracks up to 0.2 millimeters, so I think we should check whether they are clearly visible or not, and we should do so in a way that guarantees fairness. This is what we should do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. This is what we need to do. development

Vice-Chairman: Thank you very much, Mr. .
Is it all right with you, Mr. Ochiai?

Ochiai Member: Thank you, Yes. Thank you.

Vice-Chairman: Thank you very much, Mr. ? Mr. Masushima, Senior Vice-Minister for Foreign Affairs Kobayashi, please go ahead in that order.

Mr. Masashima: I have a feeling that it is over-adaptation to Thank you very much. I'm Masujima.

This is exactly the level of technology. Regarding incremental technology, it is probably like this until now, but if a better figure is obtained, it will be adopted. Instead, it is completely different. It provides great value, but the figures are not good elsewhere. I would like to know if this kind of technology will not be adopted. In other words, I understand that when you say "equivalent," you are comparing it as excellent or inferior in relation to what has been achieved by the existing value system. But are things that have great value that has not been achieved before properly evaluated separately from that? For example, from the perspective of visual inspection inspection, great value that no one had been able to approach until now has been achieved there, but for example, the image is rough, or anything is fine, but something is inferior to what has been achieved before. How does this standard deal with such things? Is it simply to apply it to the existing value system and drop it because it is inferior in comparison with it?

Vice-Chairman: Thank you very much, Mr. , Director of the Office, what do you think?

Director Wakao: Regarding the What we are looking for as inspection support technology is an inspection that is used to ensure the safety of road structures. As I said at the beginning, there was an extremely tragic accident, and in order to prevent such an accident from occurring again, it is necessary to conduct an inspection that eliminates oversight. Therefore, when conducting an inspection, it is mandatory to satisfy a certain level of performance. After satisfying that performance, as you just said, we are thinking of posting a novel initiative or something that is recognized. However, if something that should satisfy the basic performance is not recognized, I think it will be quite difficult.

Mr. Masashima: I have a feeling that it is over-adaptation to , you are simply applying existing standards and dropping them. I understand.

Vice-Chairman: Thank you very much, Mr. , thank you for waiting.

Parliamentary Senior Vice-Minister: .
I would like to ask more about the succession of this technology to the Ministry of Land, Infrastructure, Transport and

The first point is how to conduct public offering. I would like to know how to widely publicize it. We would like to take this opportunity to create opportunities for startups with new technologies and existing companies. Please tell us how you conducted public offering and what results were obtained.

My second question is, however, I believe that it inevitably costs to certify this technology. How do you cover the cost? Depending on the situation, I believe that it is necessary for each ministry to have such a budget. Please tell us what the cost is and how it is funded.

Vice-Chairman: Thank you very much, Mr. , Director of the Office, please.

Director Wakao: Regarding the public offering, the Road Bureau has made a press announcement and is soliciting public submissions so that everyone can understand it widely. As for the budget, we are using national funds and the budget of the Road Bureau.

Parliamentary Senior Vice-Minister: .

Vice-Chairman: Thank you very much, Mr. .

If you have any further questions, please send them to the Secretariat. Thank you very much, everyone. I would like to continue the proceedings because of the time.

Next, I would like to have a presentation from Mr. Takashi Hirayama, Executive Director of Road Planning, East Japan Road Management Group, Infrastructure Management Department, Infrastructure Management Division, KOKUSAI KOGYO CO.,LTD.. Thank you.

Mr. Hirayama: Thank you very much, . I would like to explain.

Now, I would like to explain about Kokusai Kogyo. It is an explanation of "development of tunnel traveling type measurement technology".

As for the content, I will explain the first point, "Introduction", and the second point, "development of hardware". The third point, "development of software". Finally, I will explain the utilization study of conventional technology and new technology.

First of all, with regard to the background of tunnel inspections, in light of the collapse of the ceiling panels of the Sasago Tunnel, which you mentioned earlier, in 2014 it was enacted into law that a close visual inspection must be carried out once every five years.

However, at about 11000 tunnels throughout Japan, inspections are conducted once every five years, and there is a Issue of shortening the construction period, reducing costs, and improving the working environment.

Regarding the tunnel inspection, as shown in Photo 1, we are currently working on a one sided alternating regulation, and at the time of inspection, we are using an aerial work platform to conduct an inspection of the nearby visual inspection. The content of the inspection is to conduct an inspection of the nearby visual inspection for cracks and such floating concrete, and to make a development drawing and a sketch drawing of the ceiling. It is a manual work.

In this regard, we are development travel-type measurement technology and inspection robots. As you just introduced, we are also publishing a three dimensional analysis system using travel-type near-infrared imaging in the Inspection Support Technology Performance Catalog specified by the Ministry of Land, Infrastructure, Transport and Tourism.

In 1998, we tried to use an HDTV camera. In the 2008 s, we tried to use a visible-light CCD camera. We could take pictures of clean tunnels, but we couldn't take good pictures of dirty tunnels.

If the tunnel is clean like this, you can take beautiful images, but we focused on whether you can take images even in a dark and sooty tunnel, so we adopted a near-infrared area sensor to development the robot.

The wavelength of near-infrared rays is larger than that of visible light. For example, apples cannot be seen with visible light, but rotten things can be seen with near-infrared rays. Or light bulbs cannot be seen with visible light, but what is inside can be seen with near-infrared rays. We have taken advantage of the fact that we can confirm the difference in characteristics.

In adopting it, we prepared a piece of concrete in a sooty state and made a test specimen with cracks on the back. We photographed this with a near-infrared camera and conducted a laboratory test. It was not visible in the digital camera image.

On the other hand, we obtained laboratory test results that cracks could not be confirmed with ordinary visible wavelengths, but cracks could be confirmed with near-infrared wavelengths. We also confirmed that wavelengths of about 850 were just right, and development the robot.

As for the robot, we have a implementation on a 4-ton truck, and it is equipped with an LED and a near-infrared camera, and we have obtained patents.

Eight cameras are installed side by side in the center. Near-infrared LED lighting is also placed around it, and for the upward facing camera, a large-capacity near-infrared LED camera is installed.

We will show the actual measurement image in a video. Since it is a traveling type, it is a situation where the image of the tunnel wall surface is taken at a speed of up to 60 kilometers per hour. Since it is near-infrared light, light and the like are invisible to the human eye, so it is possible to take images in a situation where you are just driving a truck.

At present, we have a track record of 16 tunnels throughout Japan, including several tunnels in local government, mainly tunnels of the Ministry of Land, Infrastructure, Transport and Tourism.

Next, I'm development from Software.

Regarding image analysis, we have automated the creation of a 3D model of the image taken in the near infrared. We use SfM analysis, which automatically sets the position of the camera of the image taken in this way and creates a 3D model. Therefore, we have development, which automatically creates a 3D model from the image taken in this way.

In addition, since planar development is performed from a three dimensional image, we have also automated the development from a three dimensional model to a two dimensional model.

With the flow described above, it is possible to obtain a seamless image developed in a plane like this by development the software.

This is an enlarged photo of the actual 3D model. It is shown in the video, and you can see the image from any angle, from the inside of the tunnel, and from the outside of the tunnel.

In addition, by cutting out the image, it is possible to confirm where there is a deformation, whether there are cracks, and so on. By cutting out and expanding the image in this way, I think you can see that a 0.3 mm crack with a chalk drawn at the time of inspection is confirmed at the site. In this way, we have the accuracy that a 0.3 mm crack can be confirmed on the image.

As described above, we have been creating a sketch figure and a deformation development figure by human power in this way, but we believe that by combining such image data, it is possible to obtain inspection results that are visually easier to understand and more accurate.

We have made a development of the near-infrared camera, and we are comparing it with previous high-definition cameras. If you expand the crack of 0.3 millimeters, the near-infrared can be recognized even in a dark tunnel. In the case of visible light, it is currently confirmed that there is a slight sense of out-of-focus even in a bright tunnel.

That's all about the software.

Finally, it is a comparison of the utilization of conventional technology and new technology. As an outline of the study, we considered the cost and the number of inspection days as the comparison items.

In contrast to the conventional method of using an aerial vehicle to perform proximity visual inspection and sketch, we use a traveling robot to take images to shorten the sketch.

First, the cost comparison. The comparison is based on the case where the upper stage is only one tunnel of 500 meters. The lower stage is the result of comparing the conventional type and the new technology for 10 tunnels.

As shown in this table, there are few cost advantages if we use only one tunnel. However, if we have the economies of scale of 10 tunnels, we believe that we can reduce the cost by about 30% and efficiency 50% of the on-site inspection work.

As for the source, a research project is being implemented, and this is an extract from the materials of a project conducted by industry, government, and academia.

Finally, it is a comparison of the number of inspection days.

For example, we compared the number of days required for inspections of six tunnels, from Tunnel A to Tunnel F, using conventional technologies and new technologies. It took 16 days for conventional tunnels to be inspected, but we believe that the use of new technologies as a substitute will reduce the inspection time to about eight days, a reduction of 50%.

However, if the tunnel is less than 1 kilometer, it will be no different than if it is operated by human power. If the tunnel is longer than 1 kilometer, it will be especially beneficial.

It was a rush, but that's all for your explanation.

Vice-Chairman: Thank you very much, Mr. , thank you very much.

If you have any comments or questions about the presentation, please feel free to ask, Mr. Uenoyama.

Uenoyama Member: .

I would like to ask two questions. I am sorry for the detailed discussion, but I thought that a resolution of 0.3 millimeters is amazing. How many kilometers per hour does it run, and does it actually have a spatial resolution of 0.3 millimeters?

The second is that the visual inspection has been made once every five years, and I believe that it is being judged whether it is dangerous or not. Where is the logic of the thresholds in the first place? Is it basically done only with images? Excuse me, but please tell me the details.

Vice-Chairman: Thank you very much, Mr. , what do you think?

Mr. Hirayama: Thank you very much, .

First, 0.3 millimeters has basically been demonstrated at a running speed of up to 60 kilometers. However, 1 pixel is 0.3 millimeters or more. Therefore, a scale is attached to the concrete surface, and the width of the scale is used as a reference for calculation.

These are the grounds for 0.3 millimeters.

In addition, regarding the threshold value, there is a threshold value in the inspection guidelines specified by the Ministry of Land, Infrastructure, Transport and Tourism, for example, if the crack is more than how many millimeters, it is necessary to take measures, etc., and we are using that as a reference in the inspection.

Vice-Chairman: Thank you very much, Mr. , may I speak with you?

Uenoyama Member: Yes. Thank you.

Vice-Chairman: Thank you very much, Mr. ? Mr. Masushima, please.

Mr. Masashima: I have a feeling that it is over-adaptation to Thank you very much. What I wanted to ask you is, I may have missed it, but first, I will take an image. Then, I will draw a development map, or a map. Is it correct that the creation of the map is automatically performed? In other words, if there is a crack here after looking at the image, does it mean that a person will be in the office and watch it on a monitor?

Mr. Hirayama: Thank you very much, .

Regarding the cracks that are projected now, we are using AI to challenge automatic crack extraction. However, unfortunately, at present, the extraction rate is 50% to 60%, so we are continuing deep learning with teacher data.

That's all.

Mr. Masashima: I have a feeling that it is over-adaptation to .

In relation to what we are doing under the theme of visual inspection, if we were to say that we could learn to make cracks automatically, there would be no need for people to bother looking at the images and drawing the diagrams. Moreover, if AI could do it better, I would feel that the machine would be able to see the cracks better than the people. Even in such a case, would it be necessary for people to look at the images and say validation, or would it be necessary for the technology to be recognized as being equivalent to visual inspection or would the process be unnecessary?

This may be under the jurisdiction of the Ministry of Foreign Affairs and Trade, but I believe that your company has been operating here for a long time, so could you please tell us what the outlook is?

Vice-Chairman: Thank you very much, Mr. , what do you think? I believe that this is a crucially important point for us. What are your thoughts on this?

Mr. Hirayama: Thank you very much, Unfortunately, I think it is difficult to make all of them into robots. Because, regarding whether or not a piece of concrete will fall, I don't think we can eliminate the part that we don't know until we finally hit it closely.

However, we recognize that AI will be able to extract the shape of cracks with considerable accuracy in the future, so we recognize that it is a very effective technology for screening. I'm sorry for the frustrating answer.

Mr. Masashima: I have a feeling that it is over-adaptation to .

Then, at the end, people will do the re-inspection, but the efficiency will be greatly improved, and it will be about 10 times more efficient than if people were to see everything. Will it be like this?

Mr. Hirayama: Thank you very much, Story is like that.

Mr. Masashima: I have a feeling that it is over-adaptation to .

Vice-Chairman: Thank you very much, Mr. .

How about another one?

Inadani Member: . Thank you for your very interesting explanation.

As for me, I believe that there were examples of cracks that were sooty, which overlaps with Dr. Masushima's question. Are such cases something that humans can go to the site and recognize with such certainty in the first place? I thought that you said that people are seeing things that they cannot see by using infrared rays. I thought that this is rather promising from the perspective of confirming cracks. Could you please tell us about this?

Vice-Chairman: Thank you very much, Mr. , I was also thinking so, but what is it like?

Mr. Hirayama: Thank you very much, .

We use near-infrared light, but the visual inspection close to the human eye is the most reliable. We paid attention to the fact that it may match the sooty place more than the visible light, so we made a development.

Vice-Chairman: Thank you very much, Mr. .

If you have any other questions, as I mentioned earlier, I would appreciate it if you could contact the secretariat.

Mr. Hirayama, thank you very much.

Mr. Hirayama: Thank you very much, .

Vice-Chairman: Thank you very much, Mr. , may I proceed with the agenda?

We would like to have a presentation from Mr. Shigeru Kogure, President and Representative Director of Photon Lab Co., Ltd., and Mr. Masamoto Nishikino, Chief Technology Officer of the Laser Striking Sound Division. Thank you very much.

Mr. Nishikino: In terms of . My name is Nishikino from Photon Lab. Nice to meet you.

Do you have any presentation materials?

Now, I would like to report from Nishikino, Laser Striking Sound Division of Photon Lab.

I would like to introduce it as "Laser Tapping Measurement Technology for Digital Management of Infrastructure Inspection".

Regarding this theme, we are also conducting development in the SIP project of the Cabinet Office and the project to improve the quality of road inspection related to new road technology of the Ministry of Land, Infrastructure, Transport and Tourism. I think it was the first explanation or the previous one, so there may be a little overlap, but please stay with us.

From the materials of the first Digital Extraordinary Administrative Advisory Committee Working Group, I would like to take a look at the working materials that currently classify visual inspection and on-site inspections in accordance with the compatibility of digital principles, and I would like to introduce them in accordance with the materials, so I would like to take a look at the materials.

As mentioned earlier, according to the law (Inspection Guidelines for Road Tunnels), tunnel inspection is to be carried out by checking the tunnel with one word: close visual inspection. In this, as mentioned earlier, along with the close visual inspection, which is the same as the measurement of the image, there is also a tapping inspection, which is to inspect internal defects. At the end, the bad part is knocked off. These three are called close visual inspection in set, and three are included.

In this context, led by the National Institute of Information and Communications Technology, we conducted a technical visual inspection in order to apply laser technology to remote information collection in PHASE2. Among the infrastructure maintenance, management, update, and management technologies in the first phase of SIP, we asked whether it would be possible to use a laser to remotely perform proximity development, seeing, listening to sound by hitting with a hammer, and knocking off.

At the Photon Lab, we are conducting a technical development of hitting with a laser and seeing with a laser for this second percussion sound inspection.

The purpose of this technology is to measure the microvibration of surfaces with a laser. To put it simply, the tapping sound inspection is originally the same as tapping a watermelon and listening to the difference in sound, and tapping with a hammer and listening to the difference in sound if there is a defect inside.

As an example, I didn't have a good one, so I just bought it from a vending machine. In the same way that the sound is different when the inside of a plastic bottle is clogged when it is pounded and the sound is naturally different when it is empty, I listen to the difference in the sound when the inside of concrete is floating, cracked, or hollow.

On the other hand, laser striking means hitting with a laser. What it means is, I think the video is moving to the right now, but when the laser is irradiated on the surface of the concrete and the energy of the laser is absorbed, a slightly explosive phenomenon occurs. This causes the surface to vibrate. The result of the vibration is detected by another laser, a red arrow here, which causes an explosive phenomenon, and a green laser.

What this means is that if there is a crack or cavity inside, the surface part vibrates like the membrane of a drum. When the surface vibrates, it vibrates like the right side. When an extremely high-quality laser is irradiated there, the vibration causes modulation of the frequency and wavelength of the light. By detecting this modulation, the laser detects the difference in vibration, which is basically the cause of the difference in sound heard by the hammer. This is the method of laser percussion.

In terms of what kind of data can be obtained by laser percussion sound measurement, this is converted to frequency. The horizontal axis is the frequency and the vertical axis is the intensity. In this figure, where there is a striking sound abnormality, the frequency has a peak at a little less than 4,000 hertz, and a vibration signal like "sound has sounded" appears.
This shape can be superimposed on the image on the right side, in the form of measuring several points in order against the image of the surface of the tunnel, mapping the place where there was a hitting sound abnormality in red and the normal part in green.

Again, one vibration-exciting laser vibrates the surfaces, and the vibration-measuring laser analyzes the situation. The difference in the sound heard by the ear can be corrected into digitalization, quantification, and numerical values, and finally the bad parts can be mapped.

I would like to ask to what extent this laser tapping method can be used. Of course, if you use an X-ray that looks at the human body with an X-ray, you can see deep into the tunnel lining, which is 50 centimeters long. Basically, the depth of the human tapping method is about 5 centimeters, and the tapping method is used to inspect falling objects such as falling from the surface. In terms of replacing it, it is suitable for detecting floating or peeling objects at a depth of about 5 centimeters.

So far, equipment for tunnel inspection has been installed on a 4-ton truck, and a laser development has been made. As for these equipment, as I believe it was categorized, the Ministry of Land, Infrastructure, Transport and Tourism is promoting the clarification of the application range, conditions, and implementation effects of alternative means through the Technology Catalog. It is registered in the catalog in the form of being used for concrete parts of tunnels and bridges under the non-destructive inspection of the performance catalog of the inspection support technology of the Ministry of Land, Infrastructure, Transport and Tourism.

From PHASE1 to PHASE2 of the materials, there is remote information, but PHASE2 is also evaluated by people. In short, it is about how reliable the machine we created is. It is necessary to evaluate the measurement results of our equipment for the Issue to be used for diagnostic support.

Therefore, by actually using a specimen close to various cracks such as tunnels for validation, we can quantitatively understand what kind of cracks are evaluated and how they are evaluated. Let me give you an example. We made a specimen in a shape that would crack a square block like this and pounded it from the surface. In that case, we cut each square by 1 cm and compared the results measured by laser tapping with the results pounded by two inspectors. In short, we compare what kind of judgment is made on this specimen, and compare the data, the accuracy of our equipment, and how reliable it is.

If you look at this, the reason why it is divided into red and yellow by two inspectors is that the red part is where both of them said it was defective, and the yellow part is where one of the two said it was defective. As long as people judge it, ambiguity remains in this way. Of course, it is an inspection, so it is dangerous, so in this case, the part that even one person judged to be defective is a defective area.

On the other hand, the laser striking sound is currently displayed only in green and red, but basically both of the two inspectors judge the parts that are judged to be defective as defects, including the yellow part. In addition to the tests using these test pieces, we are also comparing the inspections of actual tunnels.

This is also an image of the tunnel. There is a cable running behind it, and it is a place where the connecting part of the tunnel is often peeled off at a distance of about 10 meters each, called a joint. There is a trace of choking behind it. Regarding that part, we make a validation in the form of what kind of result can be obtained. Due to various defects of each type, a signal of abnormal hitting sound is obtained, and we compare them.

As shown in red and green, there are several signals at the moment, but the display is only abnormal or not abnormal. In order to make it more advanced in the future, as written in yellow here, we are currently conducting research and development in the second phase of SIP regarding the progress of technology such as AI. When the horizontal axis represents time and the vertical axis represents vibration, we can distinguish the vibration caused by the laser hitting sound on the surfaces, in other words, the vibration caused by defects is large, small, or absent. We take a large number of such data and use them as teacher data to perform machine learning. The classification of the actual inspection procedure is divided into four stages: normal, abnormal (small), abnormal (medium), and abnormal (large). I wrote abnormal (medium) and abnormal (large), but abnormal (large) is immediately responded to, abnormal (medium) is basically immediately responded to, and in the current tunnel, 3 and 4 are basically responded to, so only abnormal (small) to be followed up remains.

Therefore, when using it as training data, if you want to see it in the actual tunnel, there are actually only abnormal (small) defects. Therefore, in the future, abnormal (large) defects will be necessary for prevention and prediction, but this can only be done with test specimens.

In this way, we are currently evaluating the technology development related to laser striking sound with each person.

Then, in terms of how to actually operate this, it is PHASE3. As for how to bring this to automation and unmanned operation as digital data, as announced one time ago, some systems that take images of the traveling type are already running. By running this, it can be replaced with a photograph of the lining surface.

I think you had a question earlier. There are several technologies, but basically AI can extract about 80-90% of cracks. From there, some are water leakage locations where the blue part is, and cracks where the green part is. From there, more dangerous parts of cracks are estimated, and laser tapping inspection is performed.

This is why we create and store the results. Unfortunately, if you ask us if we need to use human hands in the end, we can take images, perform a tapping inspection, and even inspect cracks and internal defects. In fact, it is difficult to knock off the dust in the final visual inspection with a laser.

Technology development, such as cutting down concrete with a high-power laser, is also being carried out in a manner similar to the decommissioning of nuclear power plants. However, considering the actual inspection, it is faster for people to go and check it and knock it down with a hammer, and there is still a small amount of technology development remaining.

Now, I would like to introduce the status of the data that we have just collected. In terms of digital management, as an example, the inspection results now include a tunnel in the form of a paper called a tunnel ledger, on which we draw crack lines and take photos of what kind of cracks they are.

As I have just introduced, we will take photos of the lining image and the entire tunnel surface, and extract cracks. We will also be able to obtain the results of hammering sound. In order to change from a paper tunnel ledger to a digital one, we are currently proposing and constructing a platform aiming at a digital tunnel ledger.

In contrast to this tunnel image, the blue part is the water leakage detection point, and the red part is the crack or something like that, so I think the ideal image of robotization inspection in the future is to move from a paper ledger to a digital ledger by clicking to match the images taken from such places, and by superimposing the tapping inspection on the details of each lining image.

Finally, in terms of where we are now, I would like to introduce from this document that there are several new technologies for development, including lessons learned from the Sasago Tunnel incident.

However, new technologies use robots, and alternative technologies are very expensive, so it is not easy to replace them with people right now. In particular, it says cost, but in short, even if you can take 0.3 mm by image measurement or take sound tests at intervals of 3 centimeters, you will see all of them and they are too detailed, so it doesn't pay in terms of cost.

It just doesn't pay right now, but for example, by automating and unmanned judgment, we can accumulate digital data and connect it to future prediction and prevention, so we can contribute to the realization of a more natural society. By building a safty and security and a management method, we can contribute not only in terms of cost but also in the form of solving the shortage of human resources. That is why we are conducting research and regulation, including the Photon Lab. development.

However, on the other hand, in terms of digital management, various things have not yet been decided. For example, regarding the inspection results of the completed products that were stamped on the paper earlier, is it okay to use only electronic certification when they become digital? There are no rules about when the images were updated, or anything like that. On the other hand, the image data of the tunnel is large, and if you want to see 0.3 millimeters, it will gradually become large. Is it necessary to that extent? In fact, the images themselves will be delivered to the country or local government that requested the inspection, and there must be the rights of the person who took the images.

In addition, if data is to be disclosed to the public, if a database like the one I mentioned earlier is created throughout Japan, it will be necessary to determine whether it can be disclosed or not because it is public.

Lastly, although this is different from the above two, when I want to manage it digitally in a tunnel deep in the mountains, various radio waves are not reaching me. Of course, radio waves do not reach me in a tunnel, so I think I can finally transfer it afterwards, but when I want to do it in real time, there are some difficulties, so of course, there are good things to be said by doing it digitally, and I would like you to proceed with it, but at the same time, there are some things that need to be considered, so I would be grateful if you could consider it.

Thank you very much for taking the time to complete my presentation.

Vice-Chairman: Thank you very much, Mr. .

Do you have any questions or comments?
Mr. Masushima, please.

Mr. Masashima: I have a feeling that it is over-adaptation to Thank you very much. I learned a lot.

I would like to ask only two questions.

First of all, the solution is to use a laser. Since it is written as a tapping sound, is it correct to understand that it is a laser because we wanted to use a physical tap instead of a tap sound?

What I am asking is, if it is essential to search for defects inside, as you said earlier, for example, it is possible to find them by applying X-rays, and it seems to the untrained eye that it is essential to find the inside, and even if it is not realized by the method of tapping, there may be an approach by another technology. One question is how to think about this.

As for the other point, I would like to ask you about the part that I did not understand why you think it is not necessary to think about various things in the case of analog in the case of digital.

Mr. Nishikino: In terms of , it may be said that it is the fundamental part of laser tapping. Basically, it is written that the method of tapping and seeing with a laser was first introduced because the method of tapping and seeing with a laser should be the same in principle, although it is written that the method of tapping and seeing with a laser should be replaced as an inspection because the method of tapping and seeing with a laser is currently determined by the law of visual inspection in the vicinity of regulation.

As I was told, if it is only to see the crack, it may be possible to use an X-ray. However, if you want to see it with an X-ray, for example, if you imagine an X-ray, you need the back side where the film part is placed.

Since the back side of the tunnel cannot be seen, for example, radar waves are entered from the front and are viewed from the reflected waves. As I introduced earlier, it is used to see whether the depth of 50 centimeters of the concrete in the tunnel is 20 centimeters or 10 centimeters. In that case, even if there is a resolution of 50 centimeters or 20 centimeters, for those who see the danger of being hit by a car or a pedestrian, about 5 centimeters, which is the size of the object to be hit by the inspection, this inspection is suitable for seeing objects 5 to 10 centimeters from the surface.

We use something like ultrasonic measurement to see cracks. It can be ultrasonic, laser, or anything, but we have to attach a vibration detector to the surface. When we do so, it is a little difficult to implement in the form of remote sensing.

Therefore, in order to replace the proximity regulation that is currently performed in the legal visual inspection with robot technology, we started by applying a laser and seeing it with a laser. By applying a laser, ultrasonic waves are emitted, so naturally, we have advanced technology, and by using a beam of neutrons and the like, we can detect moisture and such things inside. However, whether it will be an inspection because it is used is a little different, and the inspection procedure does not say so, so it cannot be used there.

Mr. Masashima: I have a feeling that it is over-adaptation to will happen. Thank you very much.

Mr. Kogure: I am Kogure, the representative of , would you like to start with Kogure on the second question?

Vice-Chairman: Thank you very much, Mr. , please.

Mr. Kogure: I am Kogure, the representative of Photon Lab.

Regarding your question, the biggest problem is that in the case of infrastructure, when accidents occur, a court will be held. In the past, if drawings were stored on paper or the like, evidence would remain if they were falsified, but in the case of digital data, there is a part where evidence of falsification does not remain. For this reason, Electronic Signatures in Global and National Commerce Act and time stamps are used in the digital world, but currently, these two are not required for the management of infrastructure data.

However, in the medical care industry, where trials have already taken place in various forms, electronic medical records are kept, and in the National Tax Administration Agency, receipts are kept in digital form. Or the Patent Office's right of prior use is time-stamped to prove that it has not been tampered with, but in the case of paper, evidence is physically left, but in the case of digital, it is not left, so it has not yet been ruled as a problem.

In fact, there is no law on time stamps. It is specified by Ministerial Ordinance, so this is an excellent Japanese technology that has already become a global standard as specified by Japanese JIS. However, this is supported by blockchain technology used for Bitcoins, etc., but it is not supported by law. Therefore, Ministry of Health, Labor and Welfare and others are only providing guidance by Ministerial Ordinance.

Mr. Digital Agency, on the contrary, I would like to ask you to unify the certification of all data, for example, by legislating time stamps properly. There are few examples of Japanese technology becoming a global standard, so on the contrary, it is a technology that we can be proud of. Mr. Digital Agency, if you could summarize this, I think it would be a great promoter to use data in various industries, including infrastructure, dealing with digital.

Vice-Chairman: Thank you very much, Mr. Thank you very much for your hard assignment. President Kogure and Mr. Nishikino, thank you very much.

We will continue to advance the agenda. We would like to ask Mr. Kenji Arii, Chief Engineer of the Technology Management Department, Structural Division, CHODAI CO.,LTD., to make a presentation. Mr. Arii, nice to meet you.

Mr. Arii: CHODAI CO.,LTD.. Nice to meet you.

Then, I would like to introduce "cable tension measurement technology using vibration images".

Today, I would like to introduce the technologies listed in the Inspection Support Technology Performance Catalog compiled by the Ministry of Land, Infrastructure, Transport and Tourism.

The photograph shown here is of a type of bridge called a cable-stayed bridge over Ashida River in Fukuyama City, Hiroshima Prefecture. It is a structure in which cables are stretched diagonally from the structure called the main tower in the center to support the main girder. The cables of this cable-stayed bridge are extremely important members for the bridge, and if damage is left untreated, it will lead to the collapse of the bridge, which is an important inspection object.

A large tensile force, that is, a tensile force, is always applied to this cable. At present, confirmation of this tensile force is an inspection item to be performed in special cases. Only the appearance visual inspection is performed in the periodic inspection, but in some cases, it is performed at the same timing as the periodic inspection once every five years at the discretion of the road administrator.

This method of inspecting cable tension is a method in which measuring equipment is directly installed in close proximity to the cable, so it is difficult to quickly and safely grasp the tension on a daily basis or in the event of a disaster.

Therefore, we are working on the development of technology that enables labor saving in measurement work by DX, improvement of work safety, daily tension management, and rapid grasp of tension in the event of disasters.

The flow of tension measurement so far is shown on the left. As shown in the photo on the upper row, after installing the accelerometer, the cable is shaken manually. Then, from the installed accelerometer, the waveform data shown in the middle row can be obtained. This is called the acceleration time history waveform.

When a certain analysis is performed on this waveform, a value called the dominant frequency of the cable shown in the lower row can be obtained. Using the dominant frequency, the tensile force of the cable is estimated by inputting it into the tensile force calculation formula shown in the red frame in the center. Whether or not the estimated tensile force has a large fluctuation compared to the tensile force at the time of design is checked, and the relationship is determined.

On the other hand, the flow of this technology is shown on the right. In this technology, as shown in the upper row, the cable sway is photographed from a distance using a digital camera without approaching the cable. The photographed image is analyzed, and the displacement waveform shown in the middle row is extracted. The flow of frequency extraction, tension calculation, and safety determination after that is the same as that of the conventional technology.

We will briefly explain the principle of this technology.

The image on the left shows an enlarged view of the boundary between the cable and its background in the image taken by the digital camera. The blue dot in the figure is an arbitrary point of interest at which vibration measurement is to be performed, and the range of yellow squares around it indicates the range of interest.

On the other hand, the picture on the right shows that the point of interest moved to a certain position at the next time. In the analysis, the range in which the distribution of the brightness value of this point of interest range has the highest correlation in the image at the next time is observed, and the distance moved is obtained. By performing this analysis continuously, the point of interest is tracked one after another. Using this method, the waveform described in the previous slide can be obtained. This technology is called digital image correlation method, and it is one of the basic methods of object analysis.

Next, I will explain the results of validation on actual bridges using this technology.

This is a schematic illustration of the cable-stayed bridge that I showed you earlier. For this cable-stayed bridge cable, we compared the measurement using an accelerometer, which is a conventional method, with image measurement.

This figure shows a cross-sectional view of a cable-stayed bridge. This bridge has a total of four lanes, two up and two down, with a median strip between them, and sidewalks outside each lane. The cables are shown in red in the center. In this bridge, the bridge girder is suspended and supported by cables at the location of the median strip. An accelerometer was installed on this cable, and a camera for image measurement was installed on the sidewalk about 9 meters away from the cable.

This is the situation of the measurement. The picture on the left shows how the accelerometer is fixed to the cable with tape. On the other hand, the picture on the right shows how the camera is installed on the sidewalk and actually photographed.

This figure is a schematic illustration of the measurement position. The figure on the left is a schematic illustration of a typical cable vibration shape. The actual cable vibrates in a form in which several types of vibrations are mixed like this. We considered the shape of this vibration as much as possible, considered the position of the cable that is easier to measure, installed an accelerometer, and at the same time, determined the range to be photographed by the camera. The range shown in the red frame on the figure on the right corresponds to it.

The photos shown in the lower part here are images actually taken using image measurement. The left side is a 2K image and the right side is a 4K image, and the 4K image is higher-resolution. In each image resolution, we performed validation of the difference in measurement accuracy depending on several focus positions and the size of the focus range.

The graph on the left side shows the waveform obtained by acceleration measurement. On the right side, the relational expression between cable tension and frequency obtained by the desktop method without using the measurement results and the frequency obtained from the expression are shown in the table. A validation was performed to compare the results obtained from such acceleration waveform and relational expression with the image measurement.

This graph shows the measured waveform data decomposed for each frequency using a certain analysis method. The left side is the acceleration measurement result and the right side is the image measurement result.

The number of mountains in the graph of the image measurement on the right side is smaller than the number of mountains that can be confirmed in the graph of the acceleration measurement on the left side. This indicates that it is more difficult to measure small vibrations, which are called high-order vibrations, in the image measurement than in the acceleration measurement. However, the position of the mountain on the horizontal axis, that is, the frequency, shows almost the same value as the result of the acceleration measurement, indicating that the same value as the accelerometer side method used so far can be obtained.

This is a graph comparing the results of acceleration measurement and image measurement, with the 2K image on the left and the 4K image on the right. It shows that the same frequency as the acceleration measurement can be extracted in either case of the 2K image. Therefore, the same results as the acceleration measurement can be obtained in terms of the estimation accuracy of the cable tension as described earlier.

Please take a look at the video. This video is a visualization of the cable vibration extraction result from the image. The yellow line shows the state of the cable vibrating. It is slightly exaggerated, but you can understand how the cable vibrates when a car passes over the bridge like this.

When image measurement is performed, not only can cable tension be obtained, but also such actual shaking can be visually confirmed. It is possible to visually confirm the state of the bridge, which is difficult to understand only with numerical values, and it is possible to review the analysis as many times as necessary using recorded images. I believe that this is one of the major advantages of using this technology.

This technology, which is currently under validation, is the result of analyzing the frequency from the traffic monitoring camera images of a certain cable-stayed bridge by focusing on the cable. In this way, we believe that it will be possible to grasp the condition of the bridge by utilizing the camera images already installed on the bridge.

Finally, we summarize the future prospects of this technology.

The first is, as I introduced earlier, the monitoring of bridge conditions using images from traffic monitoring cameras and other devices already installed on bridges.

The second is that this introduction is specialized in cables, but in the future, we would like to try to validation the practicality of this technology from the perspective of understanding the condition of bridges on a daily basis or in the event of disasters by using images taken of entire bridges.

That's all for the introduction. Thank you very much.

Vice-Chairman: Thank you very much, Mr. .

Please note that some pages of Mr. CHODAI CO.,LTD.'s explanatory materials will be closed to the public in consideration of the status of coordination with related companies listed on the pages.

If you have any comments or questions about the presentation, please feel free to ask.

Ochiai Member: Thank you, * Is that all right?

Mr. Suga: We would like to hold the third meeting of the Working Group on , please.

Ochiai Member: Thank you, . Thank you for your explanation.

I believe that the content of your question is used for the inspection of bridges. When considering the physical principle, it is considered that this analysis method cannot be used for objects that move to a certain extent and objects that are static and stationary. I believe that the scope of application can be expanded a little more, and in that case, new learning will be necessary, but is it correct to assume that it will be possible to apply this method?

Mr. Arii: That's right. As you pointed out, this technology is a tracking technology for moving objects, so even if it is not an abnormal change in frequency or a short vibration, I think it is a technology that can track the position of the bridge a little or the deterioration in which displacement occurs.

Therefore, this time, it is a technology called cable tension measurement, and by applying this, I think that the displacement monitoring I mentioned earlier can be performed.

That's all.

Vice-Chairman: Thank you very much, Mr. .

Anyone else?

Parliamentary Senior Vice-Minister: : Why did you decide to start development? Did you start development because there was a public offering from the Ministry of Land, Infrastructure, Transport and Tourism? Did you start it because you thought it could be replaced by something else? Also, please tell us how much cost reduction effect you think it will have or how effective it will be.

Mr. Arii: : Actually, our company often inspects relatively long bridges, or large bridges that cross the sea, in our business. Although it is not this technology, there is a robot that inspects cables, which is also described in the development catalog. In that sense, we originally focused on cable for cable-stayed bridges, and we started to do this technology in the sense of efficient monitoring of frequency and tension, so we applied for the inspection catalog.

In terms of cost reduction effects, the current inspection method is based on accelerometers, and compared to that, efficiency or cost reduction is not so dramatic. However, as I showed in the video earlier, for example, in the sense that multiple cables can be measured at once or daily monitoring can be performed using existing cameras, I think that there is a considerable advantage in terms of efficiency, quality improvement, and work safety improvement over the conventional method of measuring by attaching an accelerometer to each cable.

Parliamentary Senior Vice-Minister: .

Vice-Chairman: Thank you very much, Mr. , are you ready? Thank you very much.

Mr. Arii, thank you very much. I was very impressed because I thought your company could build a bridge to the Bosphorus Strait.

As I mentioned earlier, some of the materials will be kept private, so we would like to ask for your understanding.

Next, I would like to have a presentation from Mr. Hitoshi Kamimoto, CEO, and Mr. Ryohei Saito, Manager, Grid Sky Way Limited Liability Business Partnership. Thank you very much.

Mr. Saito: Thank you very much for Now, Mr. Saito of Grid Sky Way will make a presentation.

Thank you very much for giving me a valuable opportunity today.

Now, I'd like to show you the PowerPoint presentation. Have you all seen it?

Then, I will tell you based on the power point.

As you can see in the upper right corner, Grid Skyway LLC is a business association consisting of two electric power companies, Tokyo Electric Power Company and Chugoku Electric Power Company, and two system vendors, NTT DATA and Hitachi, Ltd. These four companies are working on the utilization of drones in the electric power industry.

Now, I would like to explain how we would like to use drones in the electric power industry.

We have the following four use cases and ponte paintings for the purpose of responding quickly to disasters, responding to the aging and decreasing number of workers, and improving productivity. For this reason, we would like to use drones in mountainous areas where there are few people.

The first is the inspection of the power transmission equipment. As you have seen the steel tower in the mountain, we will inspect the steel tower and electric wire.

The inspection of the power transmission equipment in ② on the upper right is to inspect the steel materials of the same steel tower.

As for the third substation, in addition to the small ones near your houses, there are also very large ones, such as those four or five times the size of Tokyo Dome, and we would like to use them to patrol such places.

The last 4th is the power distribution line. This is the one standing in the form of a telephone pole near your home.

One thing I would like to explain is that we use the terms "patrol" and "inspection" separately. Patrol is a simple inspection similar to a school health checkup for humans. Regarding inspection, we use the terms "patrol" and "inspection" separately for the equipment of power transmission lines in the form of a quick and thorough look at the equipment. We use the terms "patrol" and "inspection" separately for the equipment of power transmission lines in the form of a quick and thorough look at the equipment.

In particular, during patrols and in the event of wide-area disasters, we currently use helicopters to view the situation from the sky at once, but as I will report later, even electric power companies can only manage a few helicopters, so in that sense, we would like to make use of drones.

Next is "2. Calculation example of the introduction effect of drones".

Regarding the inspection of the transmission tower as I explained earlier, the left side is a pattern in which two people climb the mountain and then climb the tower to inspect it. The right side is a pattern in which one person drives to a nearby place and then inspects the drones by autopilot.

If you look at the lower part, the green walking part and the pink ascending tower are the work to climb the steel tower. We expect that productivity will increase by about five times by compressing this area.

Next is an example of the operation of the equipment inspection system of a specific drones.

We will prepare a tablet like an iPad, and as for the drones, it will automatically approach the facility you want to see on autopilot. After that, if you press the point where you want to see here well on the tablet, it will be in focus and you can zoom further.

In this example, the steel tower has this structure as a whole, and there is a fixture called an insulator. It prevents leakage, but it is an actual example that you can zoom in to see if there are cracks here.

Next, what I just introduced is an image that the workers can see on the spot in real time. drones has an SD card in the aircraft, and you can carefully look at the images later.

What is important at that time is that, as you can see, each steel tower is very similar, so when you play it back, you don't know which steel tower you are looking at from which direction. To prevent this, we have built a system that allows you to intuitively distinguish which equipment, from which angle, and when the video was taken.

This is the high-resolution image of the inspection camera, which is higher in quality than HDTV. It is linked to when this is being photographed, but it is possible to see where drones is flying and, although it is a little difficult to see, which direction the aircraft is looking. In this figure, it is possible to see even later that the aircraft was looking at the tower called Wakatake Line No. 2 from the south.

If such an archive can be created, it will be possible to conduct AI diagnosis.

Next, here is an example of Tokyo Electric Power Company, one of our investment companies in AI diagnostics.

At TEPCO, the average age of steel towers is about 43 years old, and the oldest is 115 years old. Before rust progresses further and further, it is necessary to correct it. To do so, we take more and more still images of the steel tower, and connect them to a single photo, as shown here. After that, we cut out the background, cut off the background, and highlight only the steel tower. After that, we converted the knowledge of our workers into AI, and created an AI that makes the parts of the steel tower with strong rust stand out in red, and we are working on it.

Transmission lines have the same mechanism as steel towers, and we are working on using AI to automatically extract normal lines and broken lines that could be overlooked by human eyes.

Next, I would like to change the subject to an autopilot system.

At present, drones is often flown within a range that can be seen by the pilot using a manual control device called a propo. This is called "visual inspection flight." In visual inspection flight, the range that drones can fly is limited, and I think that it may not be possible to draw out the maximum potential.

In short, we would like to fly here. Since November 2018, the Civil Aeronautics Act has partially relaxed the regulation for watching. Until now, when flying to a place where the pilot cannot see, there was a regulation to place an assistant around here to support the visibility of the pilot. However, if certain safety requirements are met, the regulation is relaxed so that this is not necessary.

After that, it will be possible to fly to such a destination, but of course, it will fly to a place where it cannot be seen by the pilot, so it will be impossible to control it manually. Therefore, what we think is necessary is an autopilot system. If you specify in advance that you want it to fly in this order on such a power transmission line or power transmission tower, it will automatically follow you there.

This autopilot system is not completed with just one iPad like the one you just saw.

First of all, in order to specify where the drones will fly in advance, it is necessary to specify a route that will not hit any artificial objects or trees. This is a little difficult to see, but this is a picture of a steel tower that has been floated by laser surveying. What appears in the lower front is trees, although there are a few waves. First of all, it is necessary to perform laser surveying in advance and specify a place that will not hit any artificial objects or trees as a drones route or an empty road.

The second thing we need is LTE in the sky, mobile radio waves. We need to monitor and control the aircraft by constantly connecting it to the ground. With a manual device like the one you saw earlier, radio waves can only fly 1 or 2 kilometers, but with a mobile phone, we need this because it can fly anywhere.

The third and fourth is operation management. We need to monitor where other manned aircraft, helicopters, and drones are located, and the safe path in the sky, which we are confirming and highlighting, whether drones is flying away from here.

It is "8. Issue for business implementation".

With regard to the following Issue, we have been consulting with the relevant ministries and agencies. Thank you very much. I would like to introduce that if these issues are resolved, the scope of application of drones will further expand and economic efficiency will be secured.

One is to check the radio waves of mobile phones in the sky. Each mobile phone company discloses on their websites where mobile phones can be connected or not connected on the ground, but they do not currently disclose where radio waves are flying in the sky. Therefore, we are in the situation where we are flying ourselves to check whether or not mobile phone radio waves can be received.

The second is the use of mobile radio waves at altitudes of 150 meters or higher. For altitudes of 150 meters or lower, Mr. Ministry of Internal Affairs and Communications has eased the regulation, and we now have a system in place that allows the use of radio waves in the sky for communication between only these two parties, the drones user shown in this picture and the mobile phone operator. On the other hand, for altitudes of 150 meters or higher, it will take about two months more, including the processing period of Mr. Ministry of Internal Affairs and Communications.

The third is an airframe that can fly over long distances outside visual inspection. At the end of last year, a small domestic aerial filming airframe has already appeared as a result of a government project. Next, we are looking forward to the appearance of a medium-sized airframe that can fly longer distances. If possible, we would like to see a domestically produced airframe. When we communicate with domestic manufacturers, they respond faster and can convey nuance in detail.

The fourth is to improve the accuracy of flight position. We are currently loading the GPS, which is in your smartphone, on drones to grasp your position, but it is said that there are errors of up to 10 meters. Based on these errors, we are designating where to fly drones. If the receiving antenna of a quasi-zenith satellite like "Michibiki" becomes small and cheap enough to be mounted on drones, it can fly at a centimeter level, so it will be possible to take pictures from a closer distance, and we feel this problem.

I'm sorry, it's over time. In the final section, "9. Efforts toward the Air Industrial Revolution," we at Grid Skyway would like to contribute to the Air Industrial Revolution by aiming to build various elements necessary for the operation of the nationwide common route platform, the autopilot system as I just reported, starting with the inspection of power facilities.

One of the points is that, as I reported in the first part of this report, if an extremely wide-area disaster occurs, we fly helicopters, but there are only a few helicopters, so there is always a waiting list. I believe that by using drones, we can quickly understand the situation and how to respond to simultaneous disasters in a wide area.

In addition, after creating such a shipping platform, I believe it will be possible to share it with various industries, so in this example, I hope it will be a catalyst for the development of drones industries, including not only infrastructure inspections but also distribution and disaster response.

Finally, point 2 is that if we can share these routes, I believe that the overall operating cost of drones will be reduced. If we do so, infrastructure inspections and distribution will be made more drones. We believe that drones is equal to electrification, and this will contribute to carbon neutrality.

I'm sorry to have taken your time. Thank you.

Vice-Chairman: Thank you very much, Mr. , Manager, thank you very much.

As for the materials you are providing now, we would like to keep them private due to the status of coordination with related companies, as we did earlier, so please understand.

So, if you have any comments or questions about the presentation I just gave, please let me know. Mr. Ochiai, please.

Ochiai Member: Thank you, .

I believe that the use of drones is very promising. I also thought that you said that it would be better for the systems to improve so that drones can be used. What you said in particular was that the license Issue procedure of the Radio Law's Practical Application Test Station, in particular, I believe that it is over 150 meters. As is often said, I believe that there is also the Civil Aeronautics Act, but regarding flights outside visual inspection, is it correct to understand that Issue has already been resolved because the law has been revised there?

In addition, I would like to ask if there is a need for further improvement in matters related to air traffic control, which I have discussed at the regulatory reform Conference in the past, in the event that a large number of drones will be flying on autopilot.

Another possibility is to develop a database of aerial photographs taken in drones so that it can be used by everyone. I would like to ask if there is anything you have noticed that should be discussed in the future.

That's all.

Mr. Saito: Thank you very much for .

First of all, with regard to the Civil Aeronautics Act, at the end of the year before last, we asked the Civil Aviation Bureau to accept an application for measures based on a risk assessment. We have already received permission for an automatically piloted flight over 150 meters, which is called a Level 3 flight, and the system itself has been changed. We have already received permission for a Level 3 flight over 150 meters.

In addition, as you pointed out, in order to fly in the sky in the future, I believe that the more the number increases, the more necessary the operation management system will be. In this regard, the Ministry of Land, Infrastructure, Transport and Tourism, the Ministry of Economy, Trade and Industry, and the Cabinet Secretariat are currently discussing what kind of system should be adopted. We would like to propose that such a system be established from the perspective of private sector.

With regard to the development of the drones database that you mentioned, although this is my personal draft, I believe that the images taken when the drones was flown in response to the disasters do not only show the power facilities. Of course, I believe that the images also show the roads and various infrastructure around the Issue. Rather than having many airports fly and the sky becomes crowded because various business operators want to see only their own facilities, if someone can watch and share the images as a representative, I believe that there are various airports, but I believe that it will lead to the alleviation of the congestion in the sky and to the early recognition of disasters.

Vice-Chairman: Thank you very much, Mr. . You have a point.

Dr. Masujima, please.

Mr. Masashima: I have a feeling that it is over-adaptation to .

I would like to ask two questions. One is about the radio waves from the first cell phone in Issue, which I mentioned earlier. Radio waves are directional, and I think that mobile phone companies are probably trying to blow radio waves downward in order to be as efficient as possible. I don't know if they are blowing radio waves upward for the sky, or if they are in a state where they can show people how they are blowing radio waves toward the sky and say, "There are so many radio waves." I would like to ask you to tell me about this.

In addition, it seems that you are following the rules of Japan or are advancing in the direction of refining based on what you have in Japan, but if you want to fight in the world, do you think you will not be able to do this kind of thing? Because it is wonderful that you are doing this, I would like to know from the perspective of how much you can earn by doing this, and if you think that the new thing is better in total, I would like to know how you will respond.

Vice-Chairman: Thank you very much, Mr. , what do you think?

Mr. Saito: Thank you very much for First of all, regarding the mobile radio waves in the sky, as Dr. Masushima pointed out, we have received an explanation that radio waves are blowing downward. However, this does not mean that radio waves are not reaching the sky at all, so we are currently confirming that we can fly here and cannot fly here by ourselves.

Mr. Masashima: I have a feeling that it is over-adaptation to Assurance? As a mobile phone company, design is done so that radio waves blow downward, so it is OK to catch radio waves from above, but is it impossible to say that we are providing a level of service that we guarantee?

Mr. Saito: Thank you very much for This is my guess, but I think there may not be enough demand to blow radio waves upward yet.

Mr. Kamimoto: Regarding the paper on the Grid Skyway, I asked one of the mobile phone companies about the sensitivity of the upper LTE. If you pay for it, you can confirm the estimated sensitivity by simulation. However, it does not guarantee the sensitivity of radio waves, but it is just a matter of saying that it is likely to be connected by simulation, and the reality is that it costs money.

Vice-Chairman: Thank you very much, Mr. . It costs a lot.

Also, what about earning worldwide?

Mr. Masashima: I have a feeling that it is over-adaptation to , but I don't know how far it is. I understand that regulation has become like this, so it is quite necessary to go this far, and I understand well whether it is such a story or not.

Mr. Saito: Thank you very much for , if I may expand my scope a little, one of the characteristics is that helicopters do not basically approach around power transmission lines. There are tragic cases in which helicopters crashed into power transmission lines and power transmission towers, so it is relatively easy to separate manned aircraft from unmanned aircraft and drones.

If we try to virtually set the path of drones, it may be said that manned aircraft, especially helicopters, are flying in visual flight and are being operated by watching with their own eyes, so they do not know such things. This is our selfish thought, but if we can share the fact that drones passes over steel towers and power transmission lines, I expect that it will lead to the smooth division of the sky between manned aircraft and unmanned aircraft.

In addition, if we can combine the power supply system and the electric inspection system by drones while taking advantage of such characteristics for the world, I think that would be one measure. We would also share the air road we have built with other industries. As an electric utility, we are concerned that drones will spread and collide with steel towers and power transmission lines in the future. We are proceeding with this project with the intention of sharing the road we should take with everyone by taking a positive view of the current changes.

Vice-Chairman: Thank you very much, Mr. .

Dr. Inadani, I would appreciate it if you could be brief because of the time.

Inadani Member: I would like to ask you briefly about the improvement of diagnostic quality. I believe you mentioned earlier that you can find things that people overlook. In that case, flight plans are now formulated by people, but as platforms and data become more and more abundant, for example, the management of power transmission lines may be entrusted to AI.

For example, this part has not been seen recently, so let's go to see it. The AI decides the flight route, patrols, finds only dangerous parts, contacts people, and goes to fix them. Is this future vision quite possible in terms of the current speed of technology development? In other words, is there a possibility that it will lead to something that can be provided in a completely complete form?

Mr. Saito: Thank you very much for To be honest, it is still under demonstration at this point. However, on the other hand, if this can be done, things that are likely to happen now that Mr. A judges that there is no rust and Mr. B judges that there is rust will be converged, so we can renew the equipment in the same way based on the same standard. We expect that the accuracy will increase rapidly.

Vice-Chairman: Thank you very much, Mr. I guess so. There will be less and less people who can climb the steel tower.

Mr. Saito: Thank you very much for . I really do.

Vice-Chairman: Thank you very much, Mr. Secretariat and Photon Lab?

Mr. Suga: We would like to hold the third meeting of the Working Group on I would appreciate it if you could return to answer the additional questions that Dr. Ochiai asked me in the chat earlier.

Vice-Chairman: Thank you very much, Mr. Foton Lab, I'm sorry to bother you, but could you please explain?

Ochiai Member: Thank you, Should I read it out once?

Vice-Chairman: Thank you very much, Mr. That's right. Then, Mr. Ochiai, please.

Ochiai Member: Thank you, Robot. We believe that there is a possibility that the cost will decrease in the future due to factors such as mass production due to the expansion of the scope of use in the future and the advancement of production technology. Could you please tell us your outlook?

The second question is, in the case of using X-rays, will there be a safety problem even if the road is closed to unmanned aircraft and vehicles? These two questions are asked.

Vice-Chairman: Thank you very much, Mr. ?

Mr. Nishikino: In terms of Yes. I'm back.

In response to your question, I prepared the materials in a hurry, but I think there are two costs, inspection and equipment.

In terms of inspection, since robots are currently used, we would like to increase the speed of inspection. However, since robots are used, it is possible to conduct detailed inspection. Therefore, we would like to aim for a ratio of 2 to 4 times, although it is not 10 to 100 times.

In terms of cost, the cost of humans is very low, or rather, it is a strange expression, but in terms of the cost of the amount of work per day, it is far from the cost of having a robot turn it so easily. As I think it was announced earlier, the goal is to reduce the cost to two-thirds by measuring as long a distance and large an area as possible. In terms of robotization inspection that coexists with humans, I think that will be a near-future goal.

On the other hand, if we look only at the equipment, of course, it will be smaller, or it will be smaller if it is installed in a drones as you announced at the end. However, in the Japanese industry or in the mechanical development, there is a point where it can be made smaller and more expensive, and in terms of industrial structure, it can be made smaller and more convenient, so it will sell well. Therefore, I think that the cost is different between the production cost and the case of selling.

This is the outlook, or target, for the cost and speed of development.

Another question is what happens to the regulation when X-rays or neutrons are used. In the case of neutrons, the Atomic Energy regulation Agency has the RI regulation Law, so it depends on the energy. Within the range of such regulation, it is not designed to be used outdoors, so in terms of how to ensure safety, it takes a certain amount of regulation.

On the other hand, although it is not low-energy, whether we can see the back of the tunnel properly will be a different story. There are naturally radiology technicians and regulation like those in hospitals, and they are necessary. In short, the inspector has the skills of a concrete inspector, and I think that by using such a person, a different qualification will be required.

That's all.

Ochiai Member: Thank you, . I understand the problem of qualifications and regulation, but is there any particular problem with security?

Mr. Nishikino: In terms of security, it is not impossible, but in the end, regulation is a matter of security, and we are talking about whether it should not be hit or not hit too much. Basically, I don't think it will become popular and deteriorate very much, so I am imagining that it will be okay to inspect infrastructure structures.

Ochiai Member: Thank you, .

Vice-Chairman: Thank you very much, Mr. Photon Lab.

In conclusion, I would appreciate if you could give me a few words from State Minister Kobayashi.

Parliamentary Senior Vice-Minister: Thank you very much for your cooperation, all of you who participated in the second meeting of the Working Group today, all of you from private sector companies and the Ministry of Land, Infrastructure, Transport and Tourism.

As I said at the beginning, in the course of discussions and coordination with various ministries and agencies, I understand very well how much is possible with existing technology, and I feel that it is possible to see things that cannot be seen by human eyes.

In that regard, as mentioned earlier in the discussion, I think it is a point that we must consider how to evaluate the parts that cannot be done by people and the parts that can be done better.

Another point is that in terms of how to recognize such new technology and make it possible to make a implementation, the catalog method, which the people of the Ministry of Land, Infrastructure, Transport and Tourism have taken the lead in doing, is extremely effective. I felt that the existence of this method has motivated the people of private sector companies to take on new challenges and make investments, so I understand that it is an extremely effective means.

Even if the cost is the same, as Dr. Annen said, considering how to maintain infrastructure as the population decreases and the number of bearers decreases, even if the cost is the same, if we can substitute people, I believe that the current situation in Japan is that there will be great value in adopting such technology.

I believe the same thing will probably happen around the world, and I have come to realize that regulatory reform has the potential to create new growth industries that can be developed overseas.

It was a very meaningful meeting, and to be honest, the secretariat was very excited and was very excited about the interesting story. I felt that there is a possibility that a new figure will be born from private sector, in which the people of Issue and the administration will set regulatory reform together, take on new challenges, do regulatory reform, and create new industries. I think it was a very good meeting. I look forward to your continued cooperation. Thank you very much.

Vice-Chairman: Thank you very much, Mr. .

I agree with your point. A transmission line tower can be as long as 120 meters. If no one could climb it, the cost would be infinite. We are gradually entering such a world.

Thank you very much. Thank you very much for your cooperation, including the Ministry of Land, Infrastructure, Transport and Tourism, and all the presenters. You really taught me a lot of things. I think it was an extremely fulfilling conference.

Finally, I would like to ask the Secretariat to inform us of the holding of the next meeting of the Working Group.

Mr. Suga: We would like to hold the third meeting of the Working Group on at 4 p.m. today. Thank you very much.

As for today's proceedings, I believe that there are no materials other than some of the materials you pointed out that are not suitable for disclosure at the present time. Therefore, I would like to prepare the minutes later and disclose them to the public after confirming them with everyone.

In addition, if you do not have any particular objections, we would like to make today's materials available on the Digital Agency website.

Thank you very much for joining us today.

Vice-Chairman: Thank you very much, Mr. . See you at 4 p.m.