Digital Extraordinary Administrative Advisory Committee Working Group Technology-based regulatory reform Promotion Committee (4th)

Overview

Date and time: Thursday, February 9, 2023 (2023) from 3:00 p.m. to 5:00 p.m.
Location: Online
Agenda:
1. Opening
  1. Proceedings
    1. Explanation from member Ogawa
      - Considerations on Trust Collateral Measures Associated with digitalization
    2. Explanation from the Secretariat on how to proceed with the "Technology-based regulatory reform" for the time being
    3. Explanation from Mr. Suzuki
      - Efforts toward Social implementation in drones at the Fukushima RTF
    4. Explanation from the Member
      - Digital Technology Commentary Article Sample Article Writing Results validation and Evaluation Methods for Online Test Admission System (CBT) Technology
    5. Exchange of opinions
  2. Adjournment

Materials

Minutes, etc.

Date and Time

Thursday, February 9, 2023 (2023) from 3:00 p.m. to 5:00 p.m.

Location

Held online

Members present

Chairman

Hiroshi Esaki, Digital Agency Senior Expert (Architecture)

Members

Noriko Endo (Distinguished Professor, Keio University Global Research Institute)
Yusaku Okada (Professor, Department of Management Engineering, Faculty of Science and Engineering, Keio University)
Keiko Ogawa (Certified Public Accountant, Banking Capital Markets Leader LegTech Leader Partner, EY Strategy and Consulting Co., Ltd.)
Tsukasa Ogino (Representative Director of the security Council for Important Consumer Products)
KAWAHARA Yoshihiro (Professor, Graduate School of Engineering, The University of Tokyo)
Yumi Kawabata
Taro Shimada (Representative Executive Officer, President and CEO, Toshiba Corporation)
Shinji Suzuki (Designated Professor, The University of Tokyo Institute for Future Initiatives, Director of Fukushima Robot testing Field, Fukushima Innovation Coast Initiative Promotion Organization)
Takao Someya (Professor, Graduate School of Engineering, The University of Tokyo)
Keisuke Toyoda (Specially Appointed Professor, Institute of Industrial Science)
Takao Nakagaki (Professor, Faculty of Creative Science and Engineering, School of Science and Engineering, Waseda University)
Osamu Nakamura (Professor, Faculty of Environmental and Information Science, Keio University)
Ayumu Nagai (Representative Director and President of Astamuse Corporation)
Daiyu Nobori (Director of the Cyber Technology Laboratory, Information-Technology Promotion Agency)
Yutaka Matsuo (Professor, Graduate School of Engineering, The University of Tokyo)

Overview

Councilor Suga: , we would like to open the fourth "Technology-based regulatory reform Promotion Committee." Thank you for participating online this time as well.

Today's agenda consists of an explanation in the first half and an exchange of opinions in the second half. As was the case last time, we will use Webex chat. If you can comment during the explanation, please feel free to post your comments as needed. All attendees and spectators can comment on this, so we would appreciate your cooperation.
Then, I would like to start today's agenda right away. I would like to ask Chairman Ezaki to proceed with the rest of the agenda. Thank you.

Ezaki Chairman: Now, I would like to begin the proceedings.
For the fourth session, first of all, Member Ogawa will talk about "Considerations on Trust Collateral Measures Associated with digitalization" in relation to the previous agenda, and about the shortage in the previous session. After that, the Secretariat will talk about "Immediate Approaches to Technology-Based regulatory reform", and Member Suzuki will talk about "Efforts toward Social implementation of drones at the Fukushima Robot testing Field". In addition, Member Noboru will introduce technical explanation articles (prototyping). At the end, I would like to invite all members to speak freely.

First of all, I would like to ask Mr. Ogawa to explain.

Councilor Suga: We are very sorry, but there is a possibility that Mr. Ogawa's participation is delayed. If you do not mind, the Secretariat will explain first.

Ezaki Chairman: , I would like to hear an explanation from the Secretariat.

Ueki (Secretariat): Then, I would like to explain the Secretariat first. For today's Secretariat materials, I will explain the first half from Ueki and the second half from Osako. Thank you very much.

In relation to the budget that we will advance as the secretariat of Digi-in and the bills that I will explain later, we have received a wide range of opinions on Technology Map, the Technology Catalog, and other initiatives. Although there are still many Issue to be considered, we have summarized them as a way to proceed for the time being, so I would like to explain this.

Now, please turn to page 3 of the handout. First of all, I would like to talk about relations with Technology Map.

On the next page, I would like to explain about Technology Map. This is the material I presented at the Parent Committee meeting in Digital Extraordinary Administrative Advisory Committee at the end of last year. We, as the secretariat, plan to submit a package bill to promote digital regulatory reform to the ordinary session of the Diet, which is currently in session. As stated, in addition to revising regulations on paper and in-person processes in a package, which requires legal revision, we would like to take measures so that regulation can be reviewed continuously and autonomously based on the progress of technology. As circled in yellow at the bottom, based on Technology Map and other regions, we are currently considering provisions that will promote the effective use of digital technology by the competent authorities in regulation and local governments. This time, there will be a little bit of relationship between this bill and Technology Map.

On the next page, based on this situation, with regard to the ideal state of Technology Map, and in consideration of the relationship with the bill, I have organized the outline of the Technology Map to be developed by Digital Agency. The figure I have shown previously appears on the lower right again, but there has been discussion that the regulation should be generated as a dynamic one in which an arbitrary map can be obtained from the Technology Map and technology database and repository, but this time, I have organized the Digital Agency as a so-called Technology Map in the narrow sense, which I would like to make public as the of the Secretariat.

As stated in the materials, the positioning is to organize and visualization the correspondence between the types of regulations on paper and in-person processes and the technologies that can be used for the review, and based on this, we would like to aim for the use of digital technologies to advance rapidly in the regulation where the technologies can be used.

As for the components, there are some that overlap with what I have shown so far, but I have in mind the types of regulation and the specific list of regulation linked to them, the concept of digitalization and the automated phase in (ii) on the horizontal axis, and the specific information that such technology can be used in (iii). In addition, as we have discussed so far, we would like to put together the positioning of Technology Map and the relationship of responsibility, which are now listed in terms of terms of use and manuals.

Then, I would like you to go to the next page and sort out the overall picture of the efforts that are currently being made, including Technology Map, which I just explained. On the left side, I would like to include Technology Map in the narrow sense that I just explained, as Digital Agency. This is the left side.

In addition, starting with the technology catalog on the right, there are various guidelines and guidebooks that have been commented on as necessary for the utilization of other technologies. As explained later by the members, reference information for utilizing various technologies, including technical commentary articles, is taken as a whole. Regarding so-called automatically generated Technology Map, which is to extract arbitrary information from the repository of technology and regulation, if we dare to define it here, it will be positioned as an initiative to deepen the use of technology and literacy included in technical commentary articles, and it will be done in a more flexible manner. This is how I have organized it.

With regard to the specific content, how to proceed with the development, and how to make it public, as I will explain later based on the discussions at the Committee, we would like to proceed with the realization with the help of the Administration Office, which will invite applications from the public in the future, so we would like to continue to proceed with the realization while receiving the opinions of the members.

Now, I would like you to move on to the next topic, which is the overall picture, but I would like to move on to a specific review of regulation and a technical validation.

First of all, in relation to the way of proceeding with the review of regulation, this is also a document that has been submitted to Digital Extraordinary Administrative Advisory Committee at the end of this year. In the review of regulations on paper and in-person processes, a schedule has been set for approximately 10,000 provision.

On the next page, on the other hand, of the 9,669 provision Sites, there are still 1,043 validation Sites where it is difficult to see how the regulation can actually be changed without the Technology provision Site. In order to ensure that the review progresses according to the schedule, technology and validation will be necessary, so I would like to steadily advance this.

If you would like to go to the next page, as you just said, or as you said in the previous Committee discussions, you did not understand the content of the technology. Or if you are dealing with technology, you may be able to obtain suggestions that it could be used in places like regulation. That is exactly what you pointed out in the Committee. So, although it was between the end of last year and the end of the year and the beginning of the new year, we made efforts to widely solicit proposals for technology and seek opinions in Digital Agency. In a period of about one month, we received information on 347 products, services, and technology from 72 companies. First of all, as for the regulation Axis, we once mapped it and received information across the board.

On the next page, I will introduce some examples. We have received a wide range of information, including secret calculations, the fact that drones can be accessed both in the air and under water, the use of satellite data, and blockchains.

On the next page, based on the information we have received, it is like a mapping of the various information provided this time in red to the information we have been providing in Technology Map, but I believe that we have worked on the image of expanding the map and making it more concrete while receiving the information from the people of private sector, so we would like to gradually advance the review of regulation that the Secretariat has been advancing and the incorporation of information on technology.

From the next page, I will explain how to proceed with the Technology validation Project.

Osako (Secretariat): I am Osako from the Secretariat. Nice to meet you. I would like to explain how to proceed with the technical validation.

As you can see on the screen, we will secure 4.5 billion yen in the supplementary budget for fiscal 2022, and although it is a portion of that, we will promote technology validation in cooperation with the ministries and agencies responsible for regulation.

To be more specific, I explained earlier that there are approximately 1,000 provision that require technical validation. Excluding those that are individually promoted by each ministry's budget, there are slightly less than 500 provision that want to use the budget of Digital Agency. Rather than validation a little less than 500 provision one by one, we would like to implement it as efficiently as possible by classifying them according to the Issue to be solved through demonstration. provision provision

I would like to ask the following. This list summarizes the classifications. The left half is the purpose and purpose of regulation. The left major classification shows the purpose of regulation, the middle classification shows what is confirmed by inspections, etc., and the small classification shows what is actually seen for confirmation. The right half shows the types of validation cases corresponding to this classification, and they are grouped into 13 types from No. 1 to No. 13.

From the top, items 1 to 10 are inspections and checks, and whether or not it is possible to use technology instead of having people go to the site to check. In addition, regarding items 11 and 12, whether or not it is possible to mitigate the regulation of resident and dedicated personnel by using technology. And, regarding item 13, the last item, whether or not it is possible to go online for viewing is a validation.

The number of related provision on the far right is the number of validation linked to the provision types, and although there is a possibility that the number will change slightly due to future adjustments, the total number of provision is a little less than 500 as I mentioned earlier.

Since time is limited, I would like to explain the contents of several types.

Please take a look at the next page. The first is the verification of audit. As stated on the right side of the center, in regulation, which requires monitoring to understand whether there are any abnormalities in mines, facilities, and equipment, validation will be conducted to see if it is possible to use autonomous flying drones, cameras, and sensing technology as alternatives. Specifically, as stated in the main regulation targets on the lower left, we would like to conduct validation by keeping watch for dangerous goods under the area Security Act and patrolling in restricted areas under the Natural Parks Act.

Next, please skip one page and go to No. 3 on page 20. It is a demonstration of the inspection of structures. In the inspection of the management status and damage status of buildings and structures, it is to inspect whether substitution by remote execution using image analysis technology by drones and AI is possible. To be specific, I would like to advance the Issue by investigating the damage of houses and houses for the validation of Disaster Victim Certificate and inspecting the embankment of a gunpowder manufacturing facility.

Please take No. 4 on the next page. It is a demonstration of periodic inspection of equipment. Regarding confirmation of defects and degradation of equipment, we are advancing validation to see if replacement is possible by monitoring with sensors and communication devices. To be specific, we would like to advance validation by inspecting equipment in the feeding facilities of animal handling companies and gas supply equipment.

Next, please skip page 1 and go to number 6. It is a difference in the targets, but it is a field survey targeting natural objects, such as trees and cliffs in parks, rather than artificial objects. We are validation whether or not substitution is possible through the use of image analysis technology using satellite images, robots, and AI. Specifically, we would like to advance validation such as confirmation of the validity of park plans under the Natural Parks Act and the validity of conservation plans under the Natural Environment Conservation Act.

Please skip two and go to No. 9 on page 26. It is a demonstration of inspection using OCR and image analysis of drawings. Could it be possible to connect the efficiency of the completion inspection of the facility by digitizing the drawing information and using it for image analysis? There is a provision with a certain volume here, but we would like to proceed with the validation.

Next, please take No. 12 on page 29. It is a demonstration of work substitution performed by permanent or full-time employees with specific skills and experience. To what extent can technology be substituted when people at the site perform advanced work? As you can see on the lower left, we plan to advance validation in regulation, such as evaluating the condition of chicks immediately after hatching under the Poultry Farming Promotion Act by visual inspection.

Finally, I would like to ask for number 13. It is a verification of browsing. When a browsing regulation containing sensitive information is made online, a validation is made for information processing, diversion prevention technology, and identity verification technology. To be specific, we would like to proceed with a validation for browsing regulations in the processing procedure of pollution disputes.

The above is a brief explanation of the 13 types, but I would like to add three points.

With regard to my first question, although multiple regions are linked to each of these types, we would like to efficiently implement matters common to each provision by grouping them. On the other hand, since there are matters unique to each provision, we would like to closely cooperate with each provision and conduct detailed demonstrations on such unique matters.

My second question is that as a result of these demonstrations, I believe that it can be said to be performance requirements, and I would like to firmly organize these performance requirements so that they can be used as local government requirements that can be included in written specifications by each ministry, local procurement, and other procurement entities.

My third point is that the demonstration I introduced today focused on the review of visual inspection in seven items, including regulations on paper and in-person processes and resident / full-time personnel. So-called Azis and Tuby are relatively clear reviews and validation, but in parallel with these validation, I would like to firmly advance the reorganization of regulation toward further review of Issue and the examination of the necessity of demonstration toward the solution of Issue.

Finally, I would like to explain the schedule. We would like to launch an administration office for the Technology validation as early as possible in April. And for the Technology validation that can be started early, such as those for which the validation conditions are being arranged, we would like to invite business operators in June and start the demonstration around July.

With regard to Technology Map, we will prepare a specific plan by June and proceed with coordination toward the announcement of the plan no earlier than July.

We will sequentially formulate and publish a technology catalog related to Technology Map, and after the winter when the technology validation ends, we will also sequentially publish the results so that it can be used by a wide range of people concerned.

That's all the explanation from the office. Thank you.

Ezaki Chairman: .
Next, I would like to ask Member Ogawa to explain about "Considerations on Trust Collateral Measures Associated with digitalization" in relation to the previous agenda. Thank you.

Member: . Nice to meet you.
Then, I would like to share the materials. I would like to explain in about 10 minutes. Thank you very much.

It is made for your reference as "Consideration on Measures to Secure Trust Associated with digitalization".

First of all, since many stakeholders will be involved in the social implementation of Technology Map this time, I believe that your understanding and cooperation will be a condition for the social implementation. If there is no mechanism for the regulation from the perspective of each of the technology-possessing companies, the operating body, the ministries and agencies responsible for the Trust, and the technology users, it will not be possible to gain social trust. As a result, I believe that it will discourage participation and put a brake on active utilization.

Amidst this, I would like to summarize a little about accountability for securing Trust. These are actually very important themes for private sector companies these days. Today, I am preparing materials based on the example of private sector.

The point is that there is a limit, and it is very difficult to guarantee 100% that there is no risk in the introduction of new technology in terms of cost, or the speed, complexity, and expertise of the latest technology. Given that there is a limit, we believe that transparency, retroactivity, and accountability are very important.

Here are the key points from 1 to 5. First, who has what roles and responsibilities, and how to fulfill accountability for Trust. Given that risk cannot be completely eliminated, how to understand cost constraints, how to reduce risk, and to what extent to take on risk. In addition, if risk materializes, how to solve Issue at the fastest speed. In addition, I believe that it is important to understand the situation in which the internal and external environment constantly changes, and how to respond to such changes dynamically.

We believe the key is to achieve accountability, transparency, and retroactivity, as described below.

What is particularly important is that such a mechanism is not completed once it is created, but a continuous dynamic mechanism in which changes in Japan and overseas, including changes in the regulation itself, are always understood and reflected, is important as a social implementation.

I would like to talk about the trend of compliance (response to regulation) for a little reference. I am a certified public accountant, and certified public accountants naturally conduct audits on the accuracy of financial statements from the perspective of a third party. However, over this long history, events have become more complex and intense, and in addition to audits by third parties, there has been a movement to impose sworn responsibility on management for effective control development and operation.

Installation has since been widely adopted in areas such as Dodd Frank, taxation, and even anti-money laundering.

In addition, here is an example of a penalty by the US OCC. It points out the significance of the governance system itself in the first place, not only the response like a Mogura beating to improve the incompleteness of grains. I feel that the emphasis of the supervisory authority is also changing in this direction.

There is another interesting trend. Recently, we have been paying attention to "Citizen development," which can be heard on the private sector side. This is a framework in which the department closer to the front desk uses low-code and no-code tools to widely solicit DX proposals from employees in order to realize development, rather than relying only on System digitalization of the System Department. In order to properly operate this Promotion of DX framework, digital governance that considers mobility is becoming important.

Now, I'd like to share with you an example of what data / digital / technology governance is.

The key points are, first of all, the design of the governance system and the clarification of accountability. It is very important to clarify who is responsible for the Trust and what kind of responsibility they are. Next, it is necessary to clearly share what the purpose is. In addition, as I mentioned earlier, amid constraints such as costs and resources, risk cannot be eliminated. Therefore, it is important to establish a framework called the risk approach, identify and evaluate risks, and determine where to focus on. In addition, I believe that the process of monitoring whether such a mechanism is operating properly and performing dynamic and continuous validation is also essential for fulfilling accountability. Furthermore, as I mentioned earlier, a framework that can grasp and constantly reflect changing events will become the design diagram of governance of digital technology.

When considering the accountability to Trust here, one of the important themes these days is third party risk. DX is becoming a platform and an ecosystem. As a result, the risk of a third party that cannot be taken by itself will be taken in as its own risk, and it is important how to explain how to control this risk.

Next, I would like to talk a little about the risk approach. In the risk approach, while all risks cannot be eliminated, risks are firmly identified, visualization, and the significance of the impact of risks, such as impeding the achievement of objectives, is analyzed. On top of that, as shown in the lower figure on the right, even if the probability is extremely low, if the risk materializes, there are risks that must be avoided by all means, such as risks that affect human lives. On the contrary, there are risks that have extremely low social impact and will not be a big problem even if they are taken, as we have allowed so far. It is a process to confirm the policy of where to concentrate resources, such as people, goods, and money, which are both constrained.

Next, I would like to explain the types of risks. First, as shown here, we can distinguish two risks. First, there are risks inherent in the technology itself. There are risks related to the technology itself, such as the risk that the technology functions firmly as expected and the risk that the technology also has the function to control the technology to function firmly.

The second is the risk that the functional requirements themselves to achieve the purpose will be leaked. The purpose is to achieve the purpose of regulation, but if the requirements for the risks that hinder it are leaked in the first place, the purpose cannot be achieved even if the technology meets the requirements. We believe that Trust cannot be realized without responding to these two risks.

In addition, we believe that risks can be divided into three types from the perspective of who should respond to the risks. We believe that risks can be divided into three categories: risks specific to each regulation, risks common to each Technology Map category that has been classified to some extent, and risks common to all regions, for example, risks related to cybersecurity and personal data protection and risks for which there is a lack of evidence for monitoring. We believe that it is necessary to analyze these risks based on their different natures.

Also, as I mentioned in the previous meeting, I don't think it is enough to think only about technology. When new technology is introduced, the business flow that people used to do with analog will change. Even if the technology is perfect, if you don't identify and control risks in the new business process, for example, the risk of interfering with the purpose will remain no matter how well the technology functions. Therefore, I think that one key will be to consider a wide range of areas, including where the change will occur, and to identify the visualization and risk of the new business process and data flow.

Next, here is a framework called a risk control matrix. First, I said that I would take an oath to fulfill my accountability, and what I should say to fulfill my accountability, which we call an assertion, is described on the left. Next, I will identify and describe the risks that hinder it. In addition, I will describe the control activities for them. This risk control matrix is effective from the viewpoint of accountability as a framework for the sequential visualization of risk and control with assertions.

The next page shows the effectiveness of dynamic control activities. So-called discovery control, which is to take a sample and check whether there is a problem, cannot always suppress the actualization of risks. Rather, it is required to embed control technology itself, such as preventive control, as a function requirement for implementation as a function to prevent the identified risk itself in advance. In addition, if system control alone is not sufficient, analog control is also required to minimize risks.

On the next page, the risks specific to the latest technologies such as AI and RPA, which have been newly discussed recently, and third party risks such as cloud and the like are also taken up as very big themes.

In addition, in promoting paperless operations, new risks unique to digitalization will arise. New risks, such as those that can be easily replicated, are required to be identified and analyzed in accounting audits. Technology is evolving rapidly, and it is necessary to respond to these changes dynamically.

Finally, there are eight examples of "various technology Trust for securing validation." I don't have time to explain them one by one, so I will explain them one by one. First of all, I think that a mechanism such as self-evaluation by the technology-possessing company itself may be effective to a certain extent. First of all, the technology provider company will clearly take an oath about Trust to some extent.

In addition to the third party evaluation of cybersecurity, there are zero Trust evaluations, and third party evaluations called SOC reports for third party risks. Of course, there are limits to people, goods, and money, so as I explained the risk approach earlier, it is necessary to respond according to the results of effective risk analysis.

In addition, although it has been partially launched overseas, demonstration experiments with the involvement of the authorities may be effective in some cases. regulation sandboxes and hackathons with the involvement of the authorities may also be effective from the perspective of identifying risks together. In addition, what is said especially in the cybersecurity area is that fraud is a cat and mouse game. To respond to these, validation with more eyes is also effective. In some cases, validation with multiple eyes, such as the use of white hacker, is also a major trend. In any case, it is important to develop control activities that match risks.

That's all. Here's your change.

Ezaki Chairman:

Next, I would like to ask Mr. Suzuki to explain about the "Efforts toward Social implementation in drones at the Fukushima RTF."

Suzuki Member: Thank you, I am Professor Emeritus at the University of Tokyo and Specially Appointed Professor at the Future Vision Research Center. I am the director of the Fukushima Robot testing Field on a part-time basis. Today, I would like to talk about the efforts of the Fukushima Robot testing Field in about 10 minutes.

The Fukushima Robot testing Field was fully opened in 2020 as a base for the reconstruction of Fukushima. In the vast site of 1 kilometer and 500 meters in the damaged Minamisoma, various field robots, not only drones, but also underwater robots, marine robots, and ground robots, can be tested. As an example, drones is the most common.

As you are aware, a new drones was started under the Civil Aeronautics Act due to the discovery of a drones on the roof of the Prime Minister's Official Residence in 2015, and the introduction of such a regulation has expanded its use. As you may know from the news, on December 5 last year, it is expected that the new regulation will be used and utilized in the future by making it possible to conduct Level 4 flights, which are flights outside the visual inspection over people, by introducing an aircraft certification system and a national pilot license system. regulation

As mentioned earlier, as you know, it is necessary to first present major policies for the social implementation of new technologies, and it is also necessary to develop a legal regulation and a legal system for the regulation of new technologies. Utilization will advance only when practical guidelines for how to use them are prepared.

The Fukushima Robot testing Field has been active since 2019, and we are working to create various guidelines for robots centered on drones. I would like to introduce some examples. One is that drones is working on plant inspections. The guidelines of the three ministries were created by the Fire and Disaster Management Agency, Ministry of Health, Labor and Welfare, and the Ministry of Economy, Trade and Industry, and a collection of examples of their use has been issued. In terms of how to use them in practice, the Fukushima Robot testing Field has developed a practical manual and checklist from the viewpoint of practitioners, and a education curriculum for those involved in this. These are being used.

There are testing plants like this in the field, and we actually fly drones there every day. These plants are made into bridges and tunnels. They are found everywhere, but when we actually use them, we have to close roads to fly drones, or we have to stop the plant to fly drones. There is a regulation, so we use the Robot testing Field to do various kinds of validation.

This is an operation guideline for the security department, and it was developed in time for the Olympics and Paralympics. This kind of place that simulates an urban area is also made. Actual buildings and signals are made, and various demonstration experiments can be done.

In addition, as you mentioned earlier regarding risks, it is necessary to keep the risks associated with flying a drones to an acceptable level. In addition to certification of the safety of the aircraft and certification of piloting skill, it is necessary to assume in advance what risks there may be during operation and obtain permission and application in the form of a "flight manual" to confirm that the risks are within the acceptable level. However, if the way to evaluate the risks is not standardized, it is difficult to confirm them. Therefore, last year, we prepared guidelines for risk evaluation at the Fukushima Robot testing Field. We prepared these guidelines in line with the fact that Level 4 flights will be possible. Of course, it is possible that the Civil Aviation Bureau of Japan will prepare these guidelines, but because of the rapid progress of technology and the need to constantly update the content, we proposed that the Robot testing Field as a public organization be responsible for these guidelines, and we have incorporated these guidelines into the Civil Aviation Bureau's examination guidelines.

Regarding drones's risk assessments, there is already a template, and the Japanese Civil Aviation Bureau is participating in the consortium called JARUS, in which the aviation authorities of various countries participate. Based on this, the method of risk assessments is being standardized. According to this flow, the risks will be quantified, and measures to suppress them to acceptable risks will be created in this process. I don't think it is necessary to explain the content here, but in applying these international guidelines, it is necessary to conduct tailoring in accordance with, for example, Japan's legal systems and use environments. Therefore, after checking the detailed content, law is obligated to ensure that safety measures are taken for Category III flights at Level 4.

Such risk assessments are also recommended for Category II flight methods, which have been flown with permission and approval, and the Guidelines were announced in conjunction with the enforcement on December 5. You can check what they are on our website, but the Guidelines themselves, their work sheets, technical manuals for them, and what we are preparing now are samples or examples for each use cases, and we plan to make them public by the end of this fiscal year.

The use of drones is expanding, and if the use of private sector becomes dense, there is a concern that various troubles such as collisions between drones will occur, so it is necessary to manage the operation. However, it is almost impossible to manage each development by people like an aircraft, because there are so many, so we will manage them by a digital system. Research and operation of an unmanned aircraft operation management system called UTM has been conducted, and some airports are already providing services.

In Japan, such high-density operation of drones has already occurred in the event of disasters. This is actually done by monitoring the situation at the drones in the event of disasters. Helicopters also fly in such areas, so operation management is important in terms of ensuring the safety of helicopters and drones. At present, the government prohibits the flight of drones in the event of such disasters. This is for ensuring safety, but in order to operate drones and helicopters efficiently, it will be required in the future that they be used by each other under operation management. Therefore, we will not simply prohibit them, but will adjust their operation. At this stage, we have formulated guidelines that will not only be performed automatically in a system, but will be performed in an analog manner in a manual.

We are creating a education manual based on training conducted by local fire departments in Fukushima.

In order to ensure safety, it is necessary to ensure the safety of aircraft, the skills of pilots, and risk assessments during operations. In the end, it is considered necessary to conduct a third party evaluation of operators to determine whether they can implement them safely and reliably. In that sense, the Fukushima Robot testing Field is engaged in activities such as drones service quality standardization. As a result of the revision of the Japanese Industrial Standard (JIS), which has been changed from an industrial standard to an industrial standard, it has become possible to standardize service quality. Therefore, we are trying to build a mechanism to evaluate whether robot operators maintain appropriate quality and whether they can maintain and provide service quality. For operators, this will enable them to guarantee their own service quality, and for ordering parties, it will also enable them to visualization where to ask. I believe that such a mechanism will lead to the spread of drones.

In terms of future initiatives, as I mentioned earlier about aircraft safety certification, we are currently conducting activities to promote the safety of aircraft by flying actual aircraft and acquiring data. By changing this to a system that can systematically evaluate the safety from the development stage, a system that can efficiently perform the development of aircraft itself will be necessary, and I believe that a digitalization of the development process will be necessary in the future.

Of course, there is carbon neutrality, but since Fukushima is also a base for hydrogen research, the Fukushima Robot testing Field is also planning to advance efforts toward carbon neutrality in cooperation with local research facilities.

In addition, aircraft bodies are gradually becoming larger. In Europe, a small unmanned aerial vehicle is defined as one with a maximum takeoff weight of 600 kilograms. In addition, large vertical take-off and landing aircraft that can be boarded by people and flying cars are also being flight-tested at the Fukushima Robot testing Field. I believe it is necessary to further improve such an environment.

In addition, drones is not only in the air, but also in the water. There are expectations for an industrial revolution at sea, so I believe it is necessary to address this drones.

Currently, specialists from the Civil Aviation Bureau are stationed at the Fukushima Robot testing Field and work with them. For example, in the United States, in the field of aviation, the Federal Aviation Administration has transferred the rights to such an unmanned aircraft site, and the unmanned site is now able to grant flight permission. In Japan, we have not yet done so, so we need to obtain permission from the central government, but I believe that it will be necessary to transfer permission to such an testing site in the future.

That's all from me. Thank you very much.

Ezaki Chairman: , I would like to thank you for your explanation. I believe that what we want to do here, centered on drones, has been well organized and is proceeding smoothly.

Next, I would like to ask the members to explain the "Sample Article Writing Results of Commentary Articles on Digital Technology".

Member: , I'm Noboru. Around December, Mr. Deji Agency said that it would be good if there was an article that would be attractive to both the people who make decisions for each user organization and the engineers. So I was told to write one such article. I thought it would be better if I didn't say so, but I thought I would write it.

So, as a goal, it can't be helped to take time to write, so I spent about two days as a technical validation on the material I received. So I spent about five more days and worked hard for a week to write a sample that would be as good as this. Today's sample is the result of a technical validation using CBT, an online test system that was publicly invited by Digi-Cho in October. We asked about 20 companies for CBT, and one of them replied that they would like us to be a sample as soon as possible.

Originally, the tip of this was that measures against fraudulent entrance examinations were important for the Digi-cho, so when Noboru made a validation, I mostly referred to this point.
The article I wrote this time is divided into four parts. The first part is about why it is important to prevent fraud in examinations, what kinds of risks are increased compared to conventional paper, whether there are any problems other than fraud when using the Internet, and how to fix such problems.

Next, 2 is quite interesting from a technical point of view. In one online examination system I rented, when an examinee takes an examination at home using a web camera, the images are accumulated. When the AI detects that the examinee is cheating, it alerts the human proctor, so the human proctor can determine only the cheating part in a very short time. We thought we would see it from the perspective of the dishonest person and do a validation on how to cheat, so I wrote this in Chapter 2.

Chapter 3 is not about fraud this time. When you try to use this system, there may be a risk that the exam will go down or the questions will be leaked in advance. I thought I would actually try this, so I wrote down the technical validation and the results. People who make particularly important decisions sometimes want only a summary because they don't have time. So, I wrote down a summary on about 10 sheets of paper, which is about 10% of the total. I think that the decision makers, for example, in the case of computer-based exams, are the managers of public institutions and private sector institutions that are going to conduct the exam. In order for them to realize that there was such a risk, they are indifferent even if there are technical things listed, so I thought it would be better to write in a humanities and social sense.

As for the technology, I also wrote a technical article that made people feel surprised that engineers can cheat like this and that it seems to be fixed in this way.

I wanted to write this ideally, but I did it for fun while traveling for about a week, so I don't know if it has really good quality. But I think I can give a presentation for up to 15 minutes today, so I would like to show you the general structure in a few minutes.

In the first place, CBT is not the essence. The reason why I wrote this article is that when I ask another person to write a technical explanation article like this, if I do not show him a sample, I do not know how much effort I will spend and how good an article I will write, so I will make a sample. What I will say is that there are problems with CBT as a sample. I do not think that the essence of today's meeting is CBT. I think it is very interesting to recognize that CBT is interesting. I would like to discuss writing a technical article.

First of all, I would like to start with fraud. In Chapter 2, we have made a validation for fraud. In making a validation, I would like to say a little about why anti-fraud measures are important and why there are risks when fraud occurs. For example, as a matter of course, when an examination is conducted, the original purpose of selecting and evaluating people with the ability to act may be impaired if there is fraud. This is why I think the first reason why we must prevent fraud.

Second, there is a certain probability that everyone will be able to cheat. If a tool that says that it is easy to use CBT instead of paper tests, the social value of the group of successful candidates will decline, and I think that social damage will occur because the responsibility that the group of successful candidates should take in society will not be fulfilled.

Third, if the entity conducting the test is criticized as a social phenomenon such as whether it could have detected such a thing, I think that the more public a organization is, the more it will not be able to survive even if the damage is small and if it has neglected to take a few measures. I think that the manager should think that the survival of the organization is the most valuable, so I think that it must survive.

Fourth, a good person who takes the test loses money to the person who committed the fraud, and that person loses money, which causes personal damage. When I write about things like this, I remember that the purpose of our organization is to look at trees and not to look at forests, and I think that the test is like this. Then, fraud detection in the test is important. I thought he would read it to see how fraud related to fraud detection can be prevented.

First of all, regarding the application, I wrote a simple manual for normal use, or a flow. It is written like (1) to (10), and this system can be used for ordinary mark-sheet-type examinations or free-entry writing-type examinations. As a feature, as you can see on this screen, your camera appears on the upper right side of the question. And it is characterized by being monitored all the time. After this, when the test is over, for example, 60 minutes of video is sent, and it is sequentially processed by AI, and the suspicion that it is unnatural is detected. I wrote the mechanism for several pages. This is also the sales point of this product, so it was not so difficult to write this.

The actual results of the validation are as shown on this screen. When this very bad examinee is reading the past exam questions, calling and taking the exam, the AI detects it very smoothly, and the unaccuracy is 82%. If it is this much, it has already been proven, and I think it is very good because this dishonest person will be eliminated.

This is a very wonderful concept. Originally, technological progress from paper to collective computer testing occurred in the' 90s. Next, from collective to home-based types, technology for surveillance cameras spread in Oita in the late 2010s. Even so, there were limits. Finally, in 2020, technology advanced by using AI for camera surveillance. This is a wonderful point. I first described it.

However, I have raised the question of what would happen if the wonderful image of this camera could be blocked or replaced by deception. There are several ways to raise this question.

The first method is to block the camera image by pulling out the USB camera, but this does not work and can be detected properly. Then, what the fraudulent examinee thinks next is that if only the camera image part is blocked and only the test examination signal is not blocked, the camera image can be dropped. This is what everyone thinks. However, it does not work so easily. This is because there is a little technical ingenuity in this system, and the image to the camera server and the test question viewing signal are wrapped in the same code from the outside, so the fraudulent examinee cannot block the camera alone. This is a problem.

In that case, what should I do? I have written a little about the encryption called SSL, and I have written a little about the theory that a program called man-in-the-middle attack can be created in SSL. When we actually conducted the experiment, it is true that this fraudulent examinee was reading the past questions, but the result was a success. Here, SSL is an advanced concept, so it is possible to think that this is why most fraudsters cannot find it. To refute this, the first refutation is that I created this SSL program from scratch, but it took only two hours. The second refutation is that after I came up with the idea of SSL man-in-the-middle attack, I used a protocol called TCP, but it is very vulnerable to packet loss, and the length of the test browsing signal and the camera transmission signal in one communication is very different.

So, as a result of my ingenuity, if I congested the signal of my home router to the extent that it blocked packets in the upload direction of exactly 65%, it was possible to view the examination questions and to block the camera. There is such a problem.

And I wrote what the essence of this problem is. I thought there were some more ways to interfere with the camera, so I tried it. For example, there is a 100,000 pixel CMS camera like the one sold in the 2000s, and it says it is compatible with Windows98. If you connect this to the latest PC, the image quality of the web conference, which is reminiscent of 20 years ago, will come out again. And there is a very good method. When the first ID photo was taken and compared, the focus was set, and when the test started, the focus was blurred, and to my surprise, the identification was passed, the eye movement was not detected, the fraud was not found, and all of them were passed. This is the essence of this problem.

There are other things. I also mentioned that on the examinee's computer, by using the CPU separately from this app for a high load, the video was intentionally dropped, and it was successful in fraud. I also mentioned that it is possible that if the video is taken very seriously for the first 3 minutes of the exam, and for the remaining 50 minutes, the serious video is divided into 10 seconds and connected randomly, neither the AI nor the human observer can understand. I also mentioned that if you use a tab or another monitor next to it, you can cheat your eyes and get information illegally.

In addition, I mentioned a countermeasure that everything mentioned can be detected technically in this way. For example, if you open a tab in another tab, if you make a simple modification to this test system, you can detect when a tab becomes inactive. But even if it becomes inactive, isn't it impossible to prove what you were doing in the other tab? In that case, you can monitor the screen of the tester by capturing it using the capture function of this app or web browser.

However, if you use a certain API, you can create an area that cannot be captured. The question is divided into several stages and becomes more and more detailed. Since there is no cutting, it is cut down to the extent that the test implementer only needs to understand it to some extent.

In addition, how to solve some problems is also described from the perspective of Internet technology. In order to ensure availability, there are a few problems if things are left as they are, so redundancy is necessary. This method should be done in this way. This is also described. These will be greatly enjoyed by engineers, and I thought that it would be possible to make them into content that they can't help reading.

Finally, the summary describes the current situation briefly, and even those who do not have time can understand it if they read it. It took about five days to write this, and it took 2.5 days for the validation, of which 0.5 days were for writing a self-made malicious program and trying to commit fraud from the viewpoint of an attacker. If I were to ask another person to write this for about a week, the minimum content would be this much, and if I wanted to write more, I would say, well, but it can't be helped, so I would like to write this much, so I would like to submit this as a sample article in anticipation of a serious validation.

That's all for Noboru's explanation. Thank you very much.

Ezaki Chairman: . There are probably not many people who can do this much work in five days, so I have to be careful about that, but I think it is wonderful that the technology and experience are also included as a reference.

Now that you've explained it, I would like to use it as a time for exchanging opinions. I would like to hear opinions and questions from members about today's discussion, how to proceed with future discussions at the Committee, and requests for future presentation opportunities. If you would like to speak, we will designate you by raising your hand using the show of hands function. We would appreciate it if you could give us your opinions and questions.

Mr. Someya, please.

Someya Member: .
In promoting this digitalization, I understand that the talks from the teachers were centered on responses to risks and injustice.

In Mr. Noboru's speech, by disclosing more and more that injustice is done in this way and that measures are taken against it, people who want to do injustice will understand that they should do it in a sense. On the other hand, from the perspective of the person who is trying to deal with it, people who are at the forefront of trying to do injustice will think that they must go further than that.

While people who are always trying to commit fraud do not make cutting-edge and tremendous investments to commit fraud, there is also the idea that it is dangerous to disclose all of them as a measure against volume zones. On the other hand, I think there is also the idea that it can be prevented by disclosing such things. I would like to hear your view on how to handle such documents in the future.

Ezaki Chairman: , please.

Member: Thank you for your comment. That is exactly the difficult problem, so I think the answer is not to find a policy between the two extremes, but to find a middle ground depending on the content of the policy. I wrote while thinking about where the middle ground is, but I basically wrote only the content that anyone who wants to commit general fraud would come up with.

Then, you may think that this man-in-the-middle attack is difficult, but since anyone wants to manipulate the content of the communication by man-in-the-middle attack, it does not promote injustice to that extent.

However, if we proceed further, there is actually a implementation that corresponds to a specific test system for conducting man-in-the-middle attacks, and it took me about two to three hours to write it, but probably not everyone can write it, and the key will appear in the blurred part that is currently in focus. If we don't blur it, we can cheat a lot, but by blurring it, although it is certainly a fraud method, I thought that the speed at which we should take measures will be faster than the speed at which people who want to learn about SSL and cheat will appear.

Therefore, when considering the strength and weakness of the way of writing that such measures should be taken after presenting a few ways of cheating overall, we thought that we could minimize the risk that something bad would happen regardless of which way it falls in principle by determining the middle point, where the incentive to correct fraud works faster than the speed at which fraud occurs.
That's all.

Someya Member: I understand very well. Thank you very much.

Ezaki Chairman: In response, Mr. Nakamura, please.

Nakamura: Only two works by : Thank you very much.
In your response, rather than the issue of whether or not to hide information, I think it is important to consider the risks and costs as Mr. Ogawa explained earlier, and to operate from the perspective of the magnitude of the risk of something happening and the high cost of protecting against it.

I don't think there is much point in hiding information because people who notice it will notice it. Since Noboru-san is a genius, it is probably wrong to think that ordinary people cannot think of what Noboru-san is thinking, and there are people who exceed Noboru-san in the world, so in the example you explained earlier, I personally wondered how the examinee would act if he or she was reading a sentence and answering the question while reading it. In other words, if you imagine a scene in which a third party is present in a place where the camera does not capture the question, and the third party tells you the answer by reading the question aloud, if you are told that you are reading the question aloud, I think there is a greater risk that the examinee who was reading the question aloud even though he or she was not actually cheating will be treated as dishonest. I think the explanatory text is very important and very useful. If there is a person who knows the individual risk and uses it, how will you respond? We are currently conducting CBT in the School of Medicine and the School of Pharmacy, and I think that in this case, the cost of preparing and operating a special computer is commensurate with the assumed risk.

Therefore, I think it would be good if you could write down information such as how much it costs to respond to each risk in this document.
: Thank you very much.

Ezaki Chairman: : Thank you very much.
This is the same story as cybersecurity, so why information sharing and where transparency is taken will be a common Issue. We will hire a person who can handle it at the secretariat, and perhaps artificial intelligence technology can be used there.

Do you have any comments or questions from other people? I don't mind if you talk about how to proceed from now on, but in particular, the Secretariat has explained the direction it will proceed from now on. I would be very grateful if you have any good ideas, if you have to be careful about these points, or if you have any deficiencies.

Then, Mr. Ogawa, please.

Member: I'm sorry to bother you again.
I think Mr. Nakamura just said something that would be a great clue, and I think there was a talk about Azis and Toby earlier. I am also making the Risk and Control Matrix mentioned earlier in various materials this time, and I think that if you select fraud against the examination, for example, ABCD, you can listen to the fraud somewhere and show A in the back.

However, in the first place, whether the answer of ABCD is good or not, for example, if the answer of a paper that confirms the understanding very much, I think that there will be a method to confirm the degree of understanding in a different form without simply making it easy for fraud to occur.

Therefore, if what has been done so far is true, and it is said that it will be done by technology or fraud, it is very important to reconsider the limitations to some extent in the first place. This will reduce the final social cost, and if we can discuss the review of changing the way of doing things, I feel that it will be a major change.
I'll give it back to you.

Ezaki Chairman: : Thank you very much.
As I mentioned in the Secretariat, I think it is very important to pay attention to what we should do while making a distinction in common points, such as how to take measures with Azis and Tobe, and how to consider Tobe as the next step, as I mentioned in the discussion about the entrance examination.

Is there anything else from everyone?

Councilor Suga: , I think it will be difficult to comment today due to the large amount of information, so I have shared the list of technical validation projects that I just explained on the screen.

Ezaki Chairman: You are going to proceed with this from now on.

Councilor Suga: From now on, the Digi-cho will take the initiative and cooperate with each ministry and agency to participate, and we are working on the arrangement. I feel that we will be able to make a big deal.

Ezaki Chairman: are different, I feel that there are many things in common.

I also heard that another important point is that, in the explanation by the Secretariat, if the technical specifications and procurement specifications that can be used in procurement as they are are output, you can attack the problem very efficiently.

Mr. Suzuki, please.

Suzuki Member: Thank you, : Thank you very much.
I am not talking about analog or digital technologies, but for example, if there is no certification or inspection system for the data handling of business operators who conduct such inspections at the same time, the reliability of the data being handled may be lost even if the digitalization is made only with technology.

The Japanese are very meticulous about this, so they tend to trust it. However, in other countries, they are not allowed to check the data by themselves, or the data cannot be trusted unless it is checked by more than one person. In other words, they are allowed to manage the data from the standpoint that they do not trust human beings. I think that this is the most poorly understood aspect of Japanese society, and that this cannot be covered by technology alone.

There are many specific examples, but I think that these are actually important perspectives. We have just introduced drones's Service Quality JIS, and I think that it is necessary to look at how the business entity is doing these things reliably and whether it can guarantee them. So, I feel that this is an important perspective in digitalization's story, so I have talked about it from a different perspective.

Ezaki Chairman: , please.

Nakamura: Only two works by .
One is that when using technology, it is important that people who truly understand and can use the technology use it. As Mr. Suzuki said in drones, it is important that people have the technology to fly properly, and when using various technologies, it is very important to ensure the level of the experts. This is the first point. I think AI is the same.

The second point is that I think it is unclear who takes responsibility in Japanese society. If something happens, we will take responsibility at each level. As was the case with the risk earlier, there is an auditing system in management, and it has been formed as a social system that management takes responsibility in the event of a problem. For example, in security, CISO takes responsibility. I feel that unless we clarify what kind of responsibility is taken by what position and how, it will not be a system that actually works.

That's all.

Ezaki Chairman: : Thank you very much.
Well, I think that corporate governance, or rather, what to do with the governance code, must be included as very important common information.

I would also like to share one point. In particular, when we are talking about data management in the supply chain, the Ministry of Economy, Trade and Industry has been discussing the reliability of data and where it is referred to. The Ministry of Economy, Trade and Industry is also working on international standardization of data management platforms, so we need to be aware of how such data is handled as a common input.

Kawabata-san is filling it out, so Kawabata-san, would you like to talk?

Kawabata Member: Thank you, : Thank you very much.
You summarized it very well, and I originally wanted to contribute an article, but it was a rather rough time, and I left it to you. In particular, I think that it is very important to collect information in order to keep the freshness of the platform. In addition, it is very important to collect information sent by advanced experts, and I think that both the use of the platform and the freshness of the platform will be maintained. Therefore, I think that a mechanism that makes people want to contribute is necessary.

I believe that it is necessary from the perspective of design that the responsibility is very important and security is secured, but at the same time, I believe that design as an incentive is somewhat more necessary for discussions.

For example, many companies publish their own techniques, and it is said that cutting-edge engineers should write such techniques. They spend a lot of time studying them, so they think that they have to write them to some extent, but they don't make much progress. For example, I think that writing them on behalf of others will be a great incentive. I thought about whether it would be good to use this platform to draft papers.

In addition, regarding the management of such data, in Europe, there is talk of Catena-X in the manufacturing industry, and behind that, there is talk of GAIA-X, so I think it will be necessary to consider cooperation with the manufacturing industry.
That's all.

Ezaki Chairman: . We need to think about how to collect them and make a implementation.

Is there anything else?
Commissioner Ogino is writing it, so please.

Ogino Member: : I just wrote about very normal things, so in the end, cybersecurity deteriorates over time. So, aside from whether or not PDCA is good, I think it is necessary to create a mechanism that is flexible and can be updated in a timely manner so that we can always look back on the technology map.
That's all.

Ezaki Chairman: Maybe it would be better to create a part that automatically checks for updates when you create a document. It will automatically show that it has not been updated.

Ogino Member: That's right.

Ezaki Chairman: : Thank you very much.
So, Dr. Kato, how are you today? I don't have anything to write about today, but I guess you're busy.

Councilor Suga: , you may have just gotten on.

Ezaki Chairman: I see.
May I speak to Mr. Shimada?

Mr. Shimada: As for .
If you are going to do it this far, I think it is better to measure the effect. By quantifying it and showing it to everyone, motivation will be increased.

Another point is that there are about 9,000 revisions in provision, but there are only 1,000 that need to be solved by technology, so I thought that what can be done without technology should be done more quickly.
That's all.

Councilor Suga: , I would like to add that of the 9,669, technologies that have already matured, such as TV call technologies, can be adopted immediately without the need for verification or validation. In most cases, if the rules are lifted, it will be possible to go digital immediately, and we will respond immediately according to the schedule. I am aware that I am talking about the remaining 1,000 cases that you are saying cannot be taken up without confirming whether there is technology in the first place and whether it can be said that the same level of reliability can be secured for what is currently done in analog form. validation

Mr. Shimada: As for I understand.
That's why I thought it would be more appealing to Digital Agency to quantify how much tax was reduced as a result of that.

If I were to add one more thing, I would say that the fact that information can be used for digitalization instead of visual inspection means that it may be better to think again about ways to reuse information in records such as various inspection results. For example, I think there are various ways to calculate CO2 consumption from the records, but the collected data may create different values if the data are connected, so I think it is necessary to think about how to use it a little.
That's all.

Ezaki Chairman: Effect, the company will send you a notification about the management impact of a successful company, so it may be very good to share it as a success story.

Mr. Shimada: As for , when we work at a company, we are always asked what the investment effect is. So, I think it is refreshing to make a supposition that Digital Agency is also thinking about the investment effect in a cool way.

Ezaki Chairman: : Thank you very much.
May I speak to Commissioner Nakagaki?

Nakagaki Member: On page 17, which you have just shown me, what I am involved in is the facility diagnosis using sensors in No. 4 in the middle and the verification of the periodic inspection of the operating condition of the facility in No. 5. The evaluation that Mr. Shimada mentioned is an extremely important perspective, and I think it is extremely important that the changes before and after the introduction are properly quantified, because it will be used as a basis for the decision of the person who is going to introduce it.

In the inspection we are doing, we say that it used to take such a lot of man-hours and such a long time in the past, but it has been reduced to this extent. We ask for it in the form of time reduction. I think it is very easy to understand even if you look at it later.

In addition, with the original purpose of extending the life of the equipment itself and monitoring its condition, we are trying to use sensors and the like to replace those that have been inspected by people. This time, rather than the reliability of the equipment itself, the discussion has shifted to the extent to which the sensors to be introduced will last. As for parts that can be shared, if such sensors have such durability and such cases are shared, I think it will be possible to deploy them in various places.
That's all.

Ezaki Chairman: . It would be very grateful if we could find something to share and put it out well.

Mr. Matsuo, may I have your attention, please?

Member of Matsuo: : Thank you very much.
I thought what you showed me was amazing, but the secretariat asked me to write it, so what kind of request will the secretariat make next?
Ezaki Chairman: .

Member of Matsuo: , I would like to ask you about the next action of that article by Mr. Noboru, or what validation should do next. I don't want to increase my work, but I would like to ask you about your next work. What do you think about this?

Ezaki Chairman: , would you like to make an answer?

Councilor Suga: 's work is such a ridiculous thing, but. Technology Map has an image of a list of very bland words, and corporate technology is cataloged as the background of the words. I don't know what the point is even if I just show it to you. In the first place, I don't know how much I can go into it to be sure or how much I can say.

As Mr. Noboru made a presentation at the beginning, there are various people from top management of companies to people who make decisions on the ground who want to see the Technology Map and the catalog. I understand that Mr. Noboru made a validation for only one product this time, but I understand that he made a proposal that when various solutions were proposed in the type of training and test digital completion, it would be conveyed to the world in a technical article that it would be better to pay attention to these points by comparing and validation them.

However, I understand that writing such an article requires an exquisite sense of balance, including technical literacy and a mixture of hard and soft writing. I am very sorry, but I asked you to make a visualization of the image you are referring to. As a matter of course, the Secretariat does not have an image that Noboru will continue to write for free. In the future, when compiling the results of the technical validation into such an article, I would like to have the writer refer to Noboru's article as a model, and I would like to communicate that I expect this as a sample of the deliverables, although it may not be at this level.

Member of Matsuo: : Thank you very much.

As far as I remember, I understand that you made a proposal to make these articles more structured so that they can be sent out according to the purpose, or so that they can be viewed by as many people as possible. If so, I think that the content of the current PDF is wonderful, or rather, it is really amazing in such a short time, but I thought that there might be some more next steps in terms of the mechanism or what I want to do for validation, so I asked you a question.

Ezaki Chairman: articles and how to do it.

Member of Matsuo: That's right.

Ezaki Chairman: In addition, as for what we need to do next, it would be best if we could have a simulation as well as what the effect will be.

Member of Matsuo: I understand.

Councilor Suga: wrote this article. In the original proposal, I think you proposed that we need an editing system and platform that allows various people to collect information on the product from various perspectives and ensure its quality through external observation. I hope you will continue to cooperate with us in our deliberations. At the same time, rather than relying on it too much from the beginning, I would like the Digital Rincho Secretariat to take responsibility for making the various information that will come out in the future easy to see on the portal site as part of the normal Technology validation project.

Member of Matsuo: I understand. Thank you very much.

Councilor Suga: is raising his hand over this story.

Ezaki Chairman: , please.

Member: First of all, as you pointed out, I certainly argued that we would create a structured document, something like a markdown that can be versioned and can be posted on the web by more than one person. At present, it is difficult to create a document that can be versioned without development, so in fact, how we made this is based on Word PDF, which Noboru criticized earlier, saying that the administration is outrageous. Please take a look at this. On the left side of this is the table of contents of conventional words, which are also ordinary words. I had a lot of trouble, and I didn't want to write it in the same way anymore. After a while, I thought I would make it in a way that is structured, and as a result, it can be seen as PDF, PowerPoint, or HTML. If you input one source, it will automatically appear in an easy-to-read form. In order to create an incentive to write it, I thought I would make it. This is actually the hidden truth of this.

In fact, there is also a hidden intention behind this method. The reason is that Noboru is also a very incomplete technical skill, but in writing it for a while, it seems that if you stop your eyes at the part of the fraud that is interesting at first and the measures against it, you will be interested, and if you read about it next, it will be interesting to think about it more carefully after a while. I think there are three things that are interesting to anyone. The first is that the management who uses digital technology is the first, the second is that the excellent employees and employees who gather around the management are the third. The third is that the business operators who are asked to do work by the management and employees and the people who create such a system are the third. It is good to put a trick into a further version of this kind of text. The trick is that it is written on the surface about problem solving. When you try to read it more deeply, in fact, the more you read the text, the more you devise the way of writing that creates a feeling that you want to deeply understand it from the text. This is a structure of several layers hidden in one text. Even for those who do not want to study, when I thought about how to bring about the effect of naturally attracting their minds to study without being noticed, I came to the reflection that the previous sentences that we engineers have written are completely lacking in such a perspective. I regret that I have done so far in various flows, so I would like to write a sentence that includes such a hidden layer a little more from now on. I cannot write it, but I thought I would write it with such a goal for the time being, and I will submit it with a renewed determination that something that can be managed by a structured GitHub will appear in due time, or if there is none, we will create it.
That's all.

Ezaki Chairman: : Thank you very much.
In this area, natural language processing and artificial intelligence will be discussed, so Mr. Matsuo may have various opinions. Thank you very much.

You have been attending since 4 pm. Is there anything I can do for you, Mr. Endo?

Member: schedule.

I understand that the Technology validation will begin in detail, but as I have stated since last time, I would like to confirm whether the Minister in charge of Administrative Reform Okada issue will be pointed out individually or whether you intend to create a large block and make a validation within it.
Best regards

Ezaki Chairman: Secretariat respond?

Councilor Suga: At the present point in time, we are not thinking of taking out only those blocks, but I believe that we will consider them at the stage of setting performance requirements during the validation of each item.

Ezaki Chairman: That's right. However, you suggested that the point of Minister in charge of Administrative Reform Okada should be to check properly as a check point. Thank you very much.

Member: Thank you very much. I understand.

Ezaki Chairman: Then, Mr. Toyota, please.

Member: Thank you very much. I learned a lot.

To be honest, I have not organized what kind of comments I can make, but I am quite involved in the comments by Professor Matsuo and Mr. Noboru. I think that a map should be made not only in two dimensions, but also in high-dimensions and at multiple scales, so that it can be referred to and searched. If so, it is not only possible to search as characters and words, but in particular, as Mr. Ogawa said at the beginning, it is very difficult to understand the structure of a fairly large scale that cannot be referred to or imitated by conventional business or corporate models. I think that it is an area where we want to have a large reference in the contract system and the structure of various investments and recoveries. Not only the reference and detection structure embedded in advance as characters or in written form, but also something like ChatGPT is becoming a topic of conversation. What is written in characters is the background structure, and whether the structure itself is a contract, a logical structure, or a technical structure is not organized at all. However, it is not right now, but if the map can extract multiple references, it will be very suitable for Technology Map to be created in the current era, I thought, although it is my own opinion.

Ezaki Chairman: : Thank you very much.
We don't have a solution for that at the moment, so I think we will take this opportunity to accumulate it as a research development item from other places. Thank you very much.
Next, is Mr. Nagai in?

Member : Thank you very much.
I have two points that overlap with what you said earlier. First, as is the case with our company, as the amount of information increases, it becomes more difficult to cover the number of structures and dimensions. In this situation, dynamic navigation is being introduced in various ways, including the generative AI mentioned earlier. Therefore, such interactive navigation is one of the genres that have been advancing recently, and by incorporating such things experimentally, a new paradigm or search is being discussed now, and we have started an old legacy navigation. I think it would be very unfortunate if it was an afterthought after the times have changed, so I am aware that it is a little difficult to introduce, but I would like to ask you to consider such dynamic navigation. This is my first point. validation

The second point is that recently, as in the case of our company, various companies are replacing humans in monitoring and operation, and there is a new technology for validation. Aside from the technology for solving Issue itself, how to validation a large number of robots and sensors at the same time is quite costly if technology is not used. In particular, I have seen many projects that seem to have replaced humans but actually have lost much accuracy, and have entered a period of disillusionment in which they eventually return to humans. Amid this, in the current social situation where excessive expectations for digital technology are being adjusted, however, it is probably a matter of balance, and I feel that it is possible to list technologies that are used to make rational decisions to some extent on what percentage is optimal, rather than using machines to do 100% of what humans are doing. validation

Ezaki Chairman: : Thank you very much.
The people in the Secretariat were doing this by hand and using all their brains, so I have just talked with the Secretariat about how I would like to make a digitalization for this. The next place is what Mr. Nagai said, so I would like to bring it to a place where I can gain wisdom and try.
Then, Mr. Kawahara, please.

KAWAHARA, Member: , There are some parts that overlap with my previous comments, but when I saw Mr. Noboru's powerful composition, my impression is that it is a little bit of a doctor's thesis, and how to scale this is very important, and as I heard from Mr. Ogawa, I think that how to ensure the Trust of the sentences that appeared is a contradictory story, and I thought that the point of deployment is to solve this. There was already a talk about ChatGPT, and I will say it in front of Mr. Matsuo, but even if I insert a fairly appropriate sentence, if I say something like summarizing it, I will gain quite a few points, and moreover, it will summarize the points that are easy for beginners to understand, so I think it will be a good way to make a simple and orderly sentence.

However, there are also places where they answer confidently what they don't understand, so I think that there are parts that need to be carefully examined in the Trust collateral.

Ezaki Chairman: : Thank you very much.
Finally, Mr. Okada, what do you think?

Member: I'm Okada. Nice to meet you.
I think the responsibility that Professor Nakamura mentioned in the middle is important, and you have been talking about risks. In the end, the division of responsibility, the extent of responsibility, and the extent to which the collected data will be closed and opened will be linked to responsibility. Therefore, in the current flow, it is better to be open more and more. As Mr. Shimada said, if we are to measure the effect, if we are to be open more and more, various people will measure the effect. On the other hand, I think there will be people who feel uncomfortable about the responsibility of being open too much. Unless we sort out the various opinions of various people, it is not a correct answer, and I feel that there will be people who do not like either way. In particular, the more public the data is, the more likely it is to be closed even though it is open. Therefore, I think it is necessary to think about how to deal with the mental negativity of being open the more we make the collected data public.

So, while including the materials that Noboru just made, it is true that there are people who are afraid of DX in various ways because it is invisible or who dislike it, so I thought how can I show the mental hurdles and barriers of conventional analog people, not so-called IT and digital people?
That's all.

Ezaki Chairman: .
I understand that you have given us many points that we need to consider and pay attention to in the future work. Do you think you will be able to accept all of them?

Councilor Suga: . I believe that there were no particular objections to what the Secretariat explained. On the other hand, I understand that you said that we should pursue cutting-edge methods greedily, including how to accumulate and disseminate cutting-edge data, so that we do not end up with just static figures. I would like to consult with you about that, and receive advice that we should try to use this, and that we can achieve this level. I have renewed my intention.

Ezaki Chairman: That's right. In terms of budgeting, technology has to be updated all the time, so it would be easier to make the request to the regular budget, or it would be a little difficult without it. I think it would be possible to include the request not to make an amendment.

I have passed it on to you one by one, but if you have any other opinions, I still have time at the end. Is there anyone who can come? Is that okay?
Since you do not have any preference, I would like to conclude today's proceedings.

Finally, I would like to ask the secretariat to explain about the next committee meeting.

Councilor Suga: The next Committee meeting is scheduled to be held in April or later, in consideration of the budget execution I explained earlier and the Secretariat's procurement. I will contact you again.

In addition, we would like to announce today's proceedings at an early date after the Secretariat has asked you to confirm the draft minutes.

In addition, just for this time, we plan to issue a public notice of bidding for the procurement of the Secretariat in the near future, and if a business operator who is considering bidding requests it, we plan to lend the provisional version of the minutes before confirmation by everyone after receiving a written pledge on confidentiality. We would appreciate it if you could understand it.

In addition, if you do not have any particular objections to the Committee's materials today, we would like to disclose all of them on the Digital Agency website.
Thank you for attending today.

Ezaki Chairman: .
I understand that there is no objection to the presentation of the minutes of the last written pledge of confidentiality in advance.

: Thank you very much.
With that said, I would like to conclude today's meeting of the Committee. Let's meet again in April. Thank you for your hard work.